framework,version,device,op_name,kernel_source,gemm_dtype,m,n,k,latency
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,16384,20.103285047743054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,16384,40.75118679470486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,12288,33.941569010416664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,51200,62.82496473524305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,12288,15.525475396050346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,16384,84.37424045138889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,51200,136.18390570746527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,10240,33.09263780381944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,12288,61.07561577690972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,10240,13.057443406846788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,8192,20.513229370117188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,10240,50.705851236979164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,8192,10.246827019585504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,7168,17.67623053656684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,8192,43.913713243272575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,7168,9.274146185980902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,7168,35.59844292534722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,6144,19.28922356499566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,8192,29.360460069444443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,6144,30.76839531792535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,6144,8.206314086914062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,5120,14.179706149631075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,5120,6.784250471327041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,4096,12.013441297743055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,51200,254.42673068576389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,4096,14.923768785264757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,4096,5.600134955512153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,5120,25.617872450086807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,3584,10.189923604329428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,3584,4.658664703369141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,4096,20.59240044487847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,3072,8.227118598090277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,3072,13.95069800482856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,3072,4.057977040608724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,6144,22.864123874240452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,2560,8.40379163953993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,2560,8.20569101969401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,2560,6.260025872124566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,2560,3.627274831136068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,2048,3.026912053426107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,2048,6.25814946492513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,2048,5.0607350667317705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,3584,17.922861735026043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,1536,4.877752092149523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,1536,2.1209049224853516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,1536,4.591522640652127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,5120,18.594014485677082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,1024,3.311424891153971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,1024,1.4040027194552953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,1024,3.2675565083821616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,1024,3.3069992065429688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,768,2.6303314632839627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,768,1.189117325676812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,768,2.42484982808431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,512,1.9958729214138453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,512,1.9395857916937933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,512,1.1307813856336806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,512,2.0305048624674478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,256,1.1420230865478516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,256,1.7460266749064128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,256,0.9316604402330187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,256,1.4341600206163194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,128,0.9075564278496636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,65536,128,0.8688897556728787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,128,1.6823075612386067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,64,0.7577697965833875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,65536,32,0.7961555586920844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,64,1.7057626512315538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,65536,32,2.1913681030273438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,3072,11.005067613389757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,1536,4.537640889485677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,3584,13.468037923177084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,768,2.59682485792372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,65536,128,1.077622201707628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,65536,62.3283216688368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,16384,37.41307915581597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,51200,110.58000352647569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,65536,136.50306532118057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,51200,47.82147555881076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,16384,15.614342583550346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,12288,26.564821031358505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,16384,47.87987263997396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,16384,67.49436442057292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,12288,12.281223720974394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,12288,48.345682779947914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,10240,23.06057400173611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,65536,187.55653211805554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,10240,9.985838996039496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,51200,198.71469455295139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,10240,28.5860358344184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,51200,144.96352810329861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,10240,40.079125298394096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,8192,7.998898824055989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,8192,20.752621120876736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,7168,7.135325113932292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,8192,22.900770399305557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,7168,16.582383897569446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,8192,34.38037448459202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,7168,20.139295789930554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,6144,14.599158393012154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,6144,6.201876322428386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,7168,27.921727498372395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,5120,5.6105092366536455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,6144,17.379557291666668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,5120,10.944390190972221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,6144,23.976316663953995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,4096,4.3006570604112415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,5120,14.605556911892362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,4096,9.40421888563368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,65536,279.2508951822917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,5120,20.152236938476562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,4096,11.521080017089844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,3584,8.305207994249132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,4096,16.139368693033855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,3584,10.662560356987848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,3072,7.390990363226996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,3584,3.4362061818440757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,3072,3.262751897176107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,3584,14.128362867567274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,3072,10.707422892252604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,2560,2.723265753851997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,2560,6.033896128336589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,3072,8.577389187282986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,2048,4.821036444769965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,2048,1.9527573055691188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,2560,6.8867005242241754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,2560,6.653936174180772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,2048,4.944004482693143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,1536,1.3974933624267578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,1536,3.8528285556369357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,2048,4.822657691107856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,1536,3.5378907521565757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,1024,2.6020141177707248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,1024,1.0579982333713107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,1024,2.507732391357422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,1024,2.6440603468153214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,1536,3.638567182752821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,768,1.0126337475246854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,768,2.032792833116319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,768,1.9873235490587022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,512,0.7905022303263346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,768,2.0808658599853516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,512,1.527726173400879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,256,0.956909285651313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,512,1.597132470872667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,256,1.1204844580756295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,256,1.3894880082872179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,128,0.6934124628702799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,512,1.2525929345024955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,256,0.6653084225124782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,51200,128,0.6644524468315972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,64,0.542318238152398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,128,1.31621159447564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,51200,32,0.6351368692186143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,64,1.333390235900879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,51200,32,1.701802677578396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,65536,20.807907104492188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,51200,32.78490532769097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,51200,16.31158447265625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,65536,42.922685411241325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,51200,128,0.8454248640272352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,16384,11.220782809787325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,16384,5.387204488118489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,12288,8.333155314127604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,16384,12.390710618760851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,12288,3.905429416232639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,16384,21.37475077311198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,12288,9.387220594618055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,12288,15.635118272569445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,10240,3.177688810560438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,10240,7.5976808336046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,51200,66.06271023220486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,65536,48.810506184895836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,8192,2.116972393459744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,10240,7.88572014702691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,10240,13.105272928873697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,8192,6.351221296522353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,8192,6.214772542317708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,7168,5.4693785773383246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,7168,1.866620381673177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,8192,11.008355882432726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,7168,5.714304182264541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,6144,4.382250467936198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,7168,8.877115885416666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,6144,4.797208997938368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,6144,1.5415751139322917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,6144,6.331150054931641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,5120,1.2651341756184895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,5120,3.5821495056152344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,5120,3.9009441799587674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,4096,1.0188026428222656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,5120,3.731686062282986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,4096,2.976329803466797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,4096,2.9016621907552085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,4096,2.946894327799479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,3584,2.6438045501708984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,3584,0.9000444412231445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,3584,2.4267101287841797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,3584,2.5360986921522355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,3072,0.7645573086208768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,3072,2.2644897037082248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,2560,0.7541217803955078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,3072,1.9727787441677516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,2560,1.8828951517740886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,3072,2.1692960527208114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,2560,1.6541564729478624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,2048,0.5440630912780762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,2560,1.768670188056098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,1536,1.1108320024278429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,2048,1.3471564186943903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,2048,1.6607413821750219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,2048,1.4400719536675348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,1536,0.5286746554904515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,1536,1.0709048377143013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,1024,0.6922160254584419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,1536,1.1412497626410592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,1024,0.3625297811296251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,768,0.5203937954372829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,1024,0.8054693539937338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,768,0.3064684338039822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,1024,0.7810462315877279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,512,0.45530043707953555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,768,0.6360275480482314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,512,0.2726622157626682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,512,0.5004586643642849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,256,0.36214576827155215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,512,0.5068311161465114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,256,0.23054933547973633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,128,0.32592799928453237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,256,0.4597608778211806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,256,0.37604178322686094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,51200,38.16731092664931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,64,0.1753502156999376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,128,0.4203253322177463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,16384,128,0.24758667416042754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,768,0.6753493414984809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,16384,128,0.2880471017625597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,16384,32,0.27945423126220703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,64,0.4276586638556586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,32,0.48392534255981445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,16384,65536,90.05980088975694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,51200,12.371676974826388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,65536,16.21484883626302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,16384,8.476203918457031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,51200,24.087048000759548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,16384,3.8974126180013022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,51200,29.985588921440975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,16384,9.480982462565104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,16384,16.782094319661457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,65536,33.75622219509548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,12288,6.5464282565646705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,12288,2.5838470458984375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,10240,5.744474622938369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,12288,7.182550218370225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,10240,1.9140266842312281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,10240,6.0825377570258246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,12288,12.425382826063368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,10240,9.867621527777777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,8192,1.4215253194173176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,8192,4.95850584242079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,8192,4.795122782389323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,8192,6.1103261311848955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,7168,3.5797254774305554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,7168,1.3668168385823567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,7168,4.027362823486328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,51200,49.73528713650174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,7168,4.544704861111112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,6144,1.0830213758680556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,6144,3.620238410101997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,6144,3.4583928849962025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,5120,0.9394488864474826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,6144,3.3423919677734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,5120,2.778779559665256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,5120,2.694920857747396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,4096,0.7507742245992025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,5120,2.696613311767578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,4096,2.2349493238661027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,4096,2.0884026421440973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,3584,0.7007430924309624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,4096,2.18961058722602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,3584,2.2798612382676864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,3584,1.7367902331882055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,3072,0.6088373396131728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,3584,1.8967289394802516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,3072,1.8537688785129125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,3072,1.6106879976060655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,3072,1.6473119523790147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,2560,0.554308467441135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,2560,1.469739596048991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,2560,1.2939092848036025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,2560,1.3666115866767035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,2048,0.4984915521409776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,65536,37.675682915581596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,2048,1.203392876519097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,2048,1.0625680287679036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,1536,0.869401772816976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,2048,1.0712177488538954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,1536,0.4208986759185791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,1024,0.5417662196689182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,1536,0.8237262301974826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,1536,0.8076346715291342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,1024,0.3079768816630046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,768,0.5137431356641982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,1024,0.6571368641323513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,1024,0.5952817598978678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,768,0.2322791152530246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,768,0.4681430922614203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,512,0.4701697561475966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,768,0.4819679790072971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,256,0.22815023528205025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,512,0.204001784324646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,512,0.38822134335835773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,512,0.39052356614006895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,128,0.16183911429511175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,256,0.21595111158159044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,64,0.1420124504301283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,12288,128,0.18573244412740073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,128,0.3184719880421956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,256,0.5021315680609809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,12288,32,0.13729243808322483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,64,0.32567644119262695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,32,0.3526186678144667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,256,0.2926248974270291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,12288,65536,67.4178466796875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,65536,13.395645141601562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,51200,20.52422587076823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,51200,10.43747795952691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,16384,7.534906175401475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,65536,30.833367241753475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,65536,31.644093831380207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,16384,3.157712088690864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,51200,24.74337938096788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,16384,8.121785481770834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,12288,5.697557237413194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,12288,1.9659385681152344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,16384,14.366555107964409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,51200,41.39665052625868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,12288,5.977778540717231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,10240,4.592716640896268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,12288,10.02163611518012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,10240,1.5610924826727972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,10240,5.036631266276042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,8192,1.2230506473117404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,10240,7.237562391493056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,8192,3.817192925347222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,8192,3.9985440572102866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,7168,3.2114942338731556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,7168,1.0550471411810982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,8192,3.7856347825792103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,6144,0.9178596072726779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,7168,3.6430231730143228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,6144,2.428304884168837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,7168,3.2050488789876304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,12288,128,0.22031022442711723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,5120,0.7799404462178549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,6144,2.6114701165093317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,6144,2.9045492808024087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,4096,1.5975341796875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,4096,0.6125929090711806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,5120,2.6825688680013022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,5120,2.215100394354926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,5120,2.254848904079861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,3584,0.5518355369567871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,4096,1.7657768461439345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,3584,1.3418337504069011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,3584,1.5029680464002821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,3072,0.49170488781399196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,3072,1.4341999689737956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,3584,1.617400911119249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,4096,1.7110409206814234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,3072,1.2847866482204862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,2560,0.43226223521762425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,3072,1.3798364003499348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,2560,1.2747582329644096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,2560,1.0910204781426323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,2048,0.3544249004787869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,2048,0.9117404090033637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,2560,1.117669317457411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,2048,0.8641413582695855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,2048,0.9175937440660266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,1536,0.7767448955112033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,1536,0.3632168769836426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,1024,0.44993776745266384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,1536,0.6788889037238227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,768,0.35050222608778214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,1024,0.2996444437238905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,1536,0.6680302090115018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,512,0.27585776646931964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,1024,0.5238186518351237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,768,0.26870666609870064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,768,0.5533546871609157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,768,0.39855466948615176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,512,0.3330630991193983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,256,0.18442311551835802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,512,0.18514577547709146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,128,0.13443556096818712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,256,0.19524799452887642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,256,0.29646399286058217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,256,0.24103111690945098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,512,0.40279022852579754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,64,0.11183910899692112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,1024,0.5014186435275608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,10240,128,0.16144622696770564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,128,0.27076623174879283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,10240,32,0.1183004379272461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,64,0.2760106722513835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,32,0.28680798742506236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,10240,65536,57.432047526041664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,65536,10.990239461263021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,51200,16.567774454752605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,51200,8.478519015842014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,16384,5.8011902703179254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,65536,24.014292399088543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,51200,19.99561225043403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,16384,2.2249279022216797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,16384,6.469000922309028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,12288,4.2861785888671875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,51200,29.281178792317707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,12288,1.5248222351074219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,16384,10.906525505913628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,12288,6.092342376708984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,10240,1.2606915367974174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,12288,4.798367818196614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,10240,3.6132604810926647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,10240,3.8523165384928384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,10240,4.480610741509332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,8192,0.9224000506930881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,8192,3.1985556284586587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,8192,3.0838951534695096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,8192,2.9149769677056203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,7168,2.8185848659939237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,7168,0.8106897672017416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,7168,2.56876352098253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,7168,2.5199190775553384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,6144,0.7079777717590332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,6144,2.1611476474338107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,6144,2.200177722507053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,6144,2.1619955698649087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,5120,1.877810584174262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,5120,0.59952179590861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,5120,1.7989813486735027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,4096,1.4659324222140844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,5120,1.6887075636121962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,4096,0.492428461710612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,4096,1.4768053690592449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,4096,1.423167970445421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,3584,1.1612408955891926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,3584,0.4418666627671983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,3584,1.2193920347425673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,3584,1.2624293433295357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,3072,1.1559893290201824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,3072,0.38220445315043133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,3072,1.0775022506713867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,65536,26.248701307508682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,3072,1.0910719765557182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,2560,0.8651270866394043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,2560,0.3320151170094808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,2560,0.8909155527750651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,2048,0.2863333225250244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,2048,0.7183626492818197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,2560,0.8980941772460938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,2048,0.7392195595635308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,1536,0.5305760171678331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,2048,0.8916195763481988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,1536,0.24687555101182726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,1024,0.3572337892320421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,1536,0.5725928942362467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,1024,0.18631199995676676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,768,0.2919982274373372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,1024,0.4711288876003689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,512,0.21071467134687635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,1024,0.41905866728888613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,768,0.20605866114298502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,768,0.3226693206363254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,256,0.13495200210147434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,768,0.3453528881072998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,512,0.29197777642144096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,512,0.2710631158616808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,128,0.10881244473987156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,256,0.2326000001695421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,256,0.14246133963267008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,256,0.19734044869740805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,1536,0.5944337844848633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,64,0.08803200059466892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,128,0.13862133026123047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,8192,32,0.09235466851128472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,128,0.2159013350804647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,65536,44.643395317925346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,64,0.21977421972486708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,8192,32,0.22877600457933214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,8192,512,0.1755653354856703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,10240,128,0.18614578247070312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,65536,9.749560885959202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,51200,15.053916083441841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,65536,17.99728054470486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,16384,4.48910649617513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,51200,7.492875840928819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,51200,18.37563747829861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,16384,1.7163893381754558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,12288,3.737175835503472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,16384,7.660298665364583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,16384,5.739972432454427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,12288,1.2899235619439018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,12288,4.622200012207031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,12288,4.187511867947049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,10240,3.298512988620334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,51200,29.53048366970486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,10240,1.093543158637153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,10240,3.3674924638536243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,10240,3.4185742272271047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,8192,0.795838197072347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,8192,2.7081360287136502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,8192,2.648729748196072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,8192,2.6249023013644748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,7168,2.36464712354872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,7168,0.7391502062479655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,7168,2.2477874755859375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,7168,2.1929876539442277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,6144,2.067826588948568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,6144,0.6161555714077419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,6144,1.950842751397027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,6144,1.864007102118598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,5120,1.6304399702284071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,5120,0.5223840077718099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,5120,1.6008115344577367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,5120,1.702590200636122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,4096,1.3218417697482638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,4096,0.4361226558685303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,4096,1.2418684429592557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,4096,1.4676791297064886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,3584,1.086658689710829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,3584,0.3950231075286865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,3584,1.0781129201253254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,3584,1.1483724382188585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,65536,38.780602349175346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,3072,0.3516799873775906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,3072,0.9169662263658312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,3072,0.9150515662299262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,2560,0.8323120011223687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,2560,0.3074302143520779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,2560,0.7948168648613824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,2560,0.7599262131585015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,2048,0.6233626471625434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,2048,0.2623315652211507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,1536,0.46999645233154297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,2048,0.6231999927096897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,2048,0.7845422426859537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,1536,0.2719004419114855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,1536,0.6186800003051758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,1024,0.30737421247694224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,1536,0.4932311375935872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,1024,0.3711377779642741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,1024,0.16978489028082955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,3072,0.9602720472547742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,768,0.1590835518307156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,768,0.29115910000271267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,1024,0.5063964525858561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,768,0.29411111937628853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,512,0.17834666040208605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,256,0.12888444794548884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,512,0.23947376675075957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,512,0.23736622598436144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,128,0.0952755543920729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,256,0.12666666507720947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,256,0.204584002494812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,768,0.24564178784688315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,256,0.17853066656324598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,64,0.07625689109166463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,7168,128,0.12571733527713352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,128,0.19029066297743055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,32,0.08053244484795465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,7168,512,0.1750426689783732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,64,0.1949991120232476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,7168,32,0.19951289229922822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,65536,24.282457139756943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,8192,128,0.15187022421095106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,65536,8.477481418185764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,51200,12.019368489583334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,51200,6.3576143052842875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,65536,15.650183783637154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,16384,4.308893415662977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,16384,1.4220932854546442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,51200,15.560402764214409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,65536,20.003780788845486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,16384,4.958923763699002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,12288,3.4294649759928384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,16384,6.516121758355035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,12288,1.1121821933322482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,12288,3.7433111402723522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,10240,0.8971563975016276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,12288,3.519284354315864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,10240,2.9510826534695096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,10240,2.737566206190321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,10240,2.908807118733724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,8192,0.7036044332716199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,8192,2.2530434926350913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,8192,2.1857503255208335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,51200,24.391425238715275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,8192,2.2559483846028647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,7168,2.067899492051866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,6144,1.6760311126708984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,7168,0.6036835776435004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,7168,1.9169332716200087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,6144,1.6238462660047743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,6144,0.6555280155605739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,65536,29.83852810329861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,5120,0.5086862246195475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,7168,1.9441093868679469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,5120,1.3188222249348958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,6144,1.6841964721679688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,5120,1.296079953511556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,4096,0.46456178029378253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,4096,1.1092728508843315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,5120,1.3719636069403753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,3584,0.9179582595825195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,3584,0.3967599868774414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,4096,1.0957298278808594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,4096,1.135567135281033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,3584,0.9311190711127387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,3072,0.8194586435953776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,3072,0.30914399358961314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,3584,0.9395501878526477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,2560,0.6926808887057834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,2560,0.27508799235026044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,3072,0.9176710976494683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,3072,0.8553262286716037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,2048,0.5097466574774848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,2560,0.678735997941759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,2048,0.23582222726609972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,2560,0.7322755389743381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,1536,0.40335644616021055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,1536,0.19071288903554282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,2048,0.6150782373216417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,2048,0.5522240002950033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,1024,0.27113244268629283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,1536,0.429916434817844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,1536,0.6195199754503038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,768,0.20838844776153564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,1024,0.1570604509777493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,512,0.15889422098795572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,768,0.16131644778781468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,1024,0.49168978797064883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,768,0.28656087981330025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,1024,0.3235315481821696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,256,0.11751377582550049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,768,0.27496356434292263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,512,0.20574043856726754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,512,0.14555288685692683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,128,0.08452533351050483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,256,0.11910133891635472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,256,0.18172978030310738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,256,0.15330399407280818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,64,0.07020178106096056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,6144,128,0.11535822020636664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,6144,32,0.07077688641018338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,128,0.16462488969167074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,128,0.11910933918423122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,64,0.1671919955147637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,6144,32,0.17167466216617158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,6144,512,0.2627591027153863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,65536,6.664575788709853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,51200,10.621419270833334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,51200,4.965040842692058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,16384,3.9358168707953562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,51200,13.324001736111112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,7168,128,0.13547200626797146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,65536,15.393320719401041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,16384,1.2111555735270183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,16384,4.495828416612413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,12288,2.6207154591878257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,65536,17.36895582411024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,12288,0.8847466574774848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,12288,3.0140177408854165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,16384,4.113748338487413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,10240,0.7542817857530383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,12288,2.9414694044325085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,10240,2.4827520582411027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,51200,19.80832756890191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,10240,2.3703856998019748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,10240,2.438958273993598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,8192,0.7346124649047852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,8192,2.110496097140842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,8192,1.9323378668891058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,8192,1.9516027238633897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,7168,0.5630790922376845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,7168,1.7751715977986653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,7168,1.5533075332641602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,6144,0.4667244487338596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,6144,1.5271102057562933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,7168,1.6988684336344402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,6144,1.4549261728922527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,6144,1.4456222322252061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,5120,1.2388630972968209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,5120,0.5302640067206489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,5120,1.110382186041938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,4096,0.8718461990356445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,5120,1.190643522474501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,4096,0.3774240016937256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,4096,0.9007831149631076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,3584,0.3150204552544488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,3584,0.8063857820298937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,3072,0.7116719881693522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,3584,1.0470231374104817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,3584,0.7804533110724555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,3072,0.7032195727030436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,3072,0.3149955537584093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,2560,0.5224044587877061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,4096,0.8970799975925021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,3072,0.9240621990627713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,2560,0.26212623384263783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,2560,0.5942026774088541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,2048,0.43520000245836044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,2048,0.20416622691684297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,1536,0.3257582187652588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,2560,0.5785848829481337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,2048,0.4671102099948459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,1536,0.16593689388699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,1536,0.3650711112552219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,1024,0.22665778795878092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,2048,0.48816797468397355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,768,0.1891253391901652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,1536,0.4244675636291504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,1024,0.13596267170376247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,1024,0.2846826712290446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,512,0.12875200642479792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,768,0.126674665345086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,768,0.2789439890119764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,1024,0.27835554546780056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,256,0.09346755345662434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,768,0.22758844163682726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,512,0.12053244643741184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,512,0.17425866921742758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,128,0.07095199823379517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,512,0.18162843916151258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,256,0.10240088568793403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,64,0.056201775868733726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,256,0.1615262164009942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,256,0.13170044951968723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,5120,32,0.0604648921224806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,5120,128,0.10172444581985474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,65536,27.728071424696182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,5120,128,0.10130577617221409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,64,0.1394453313615587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,32,0.14320977528889975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,5120,128,0.14113067256079778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,51200,3.792857699924045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,65536,5.120959811740451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,16384,2.813530604044596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,51200,9.41575453016493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,16384,3.538720024956597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,65536,10.298099941677517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,16384,0.9773289362589518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,51200,10.988928053114149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,12288,2.1582347022162542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,16384,3.3255333370632596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,12288,0.739009804195828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,12288,2.2965325249565973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,12288,2.396846135457357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,10240,1.9494569566514757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,10240,0.6435573365953233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,10240,1.986996544731988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,10240,1.9344488779703777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,65536,14.40785386827257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,8192,1.4806266360812719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,8192,0.4958800209893121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,51200,14.828638712565104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,8192,1.6114755206637914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,7168,0.4899297820197211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,8192,1.7854294247097438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,7168,1.3870337804158528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,6144,1.2554089228312175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,6144,0.39328355259365505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,7168,1.362844467163086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,7168,1.3579093085394964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,5120,0.8810693422953287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,5120,0.3364773326449924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,6144,1.1661128997802734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,4096,0.6962569024827746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,4096,0.2753102249569363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,5120,0.9819200303819445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,5120,1.1168897416856554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,4096,0.8488791253831652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,3584,0.2524160014258491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,4096,0.8124506738450793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,3584,0.6335751215616862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,6144,1.2048426734076607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,3584,0.6696497599283854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,3584,0.6759582095675998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,3072,0.4632559882269965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,3072,0.23186845249599883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,2560,0.4330577850341797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,3072,0.5581875377231175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,3072,0.5948799981011285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,2048,0.3198435571458605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,2560,0.48041868209838867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,2560,0.4975688722398546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,2048,0.20746755599975586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,1536,0.2637760109371609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,2048,0.43092089229159886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,1024,0.18413421842787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,1536,0.32349689801534015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,1536,0.3974951108296712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,1024,0.23676623238457573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,2560,0.24082488483852812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,1024,0.11789155006408691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,1024,0.23748087882995605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,768,0.11200888951619466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,768,0.17628267076280382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,512,0.10655022329754299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,768,0.19033066431681314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,2048,0.4042088985443115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,256,0.07363555828730266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,512,0.14342933230929905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,512,0.10021422306696574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,512,0.15004977915022108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,128,0.0554693341255188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,256,0.12248799535963271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,256,0.10977066887749566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,768,0.1491395499971178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,64,0.04586133360862732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,128,0.11327821678585476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,128,0.08911999728944565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,4096,32,0.04751555456055535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,4096,128,0.08398844136132134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,64,0.1130666732788086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,32,0.11423288451300727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,1536,0.15227111180623373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,4096,256,0.090229332447052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,4096,65536,21.15613301595052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,65536,4.404129876030816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,51200,3.430034637451172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,16384,2.6309422387017145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,51200,8.496424357096354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,16384,0.8794533411661783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,16384,3.0020275115966797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,51200,9.91646236843533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,16384,3.025825712415907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,65536,11.796059502495659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,12288,0.697015126546224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,12288,2.0658541785346136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,12288,2.009338590833876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,12288,2.168670230441623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,10240,1.560190200805664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,10240,0.5726435449388292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,51200,13.910724216037325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,10240,1.812588373819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,8192,1.3955315483940973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,8192,0.4491644435458713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,8192,1.3751253551907008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,8192,1.422622256808811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,7168,1.2052044338650174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,10240,1.7357236014472113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,7168,0.39797245131598574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,7168,1.3232737647162545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,7168,1.2431529362996419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,6144,0.9997652901543511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,6144,1.0789830949571397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,6144,1.0668000115288627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,5120,0.7387146419949002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,5120,0.30738843811882866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,5120,0.8509502410888672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,6144,0.34736355145772296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,4096,0.6174231105380589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,5120,0.8908221986558703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,4096,0.24955911106533477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,4096,0.76327821943495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,3584,0.5432204670376247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,3584,0.22545599937438965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,4096,0.7216915554470487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,3072,0.42555822266472715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,3584,0.8617093298170301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,3584,0.6129128668043349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,3072,0.21717245048946807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,3072,0.5389484299553765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,2560,0.4047164387173123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,2560,0.1827546623018053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,3072,0.528656005859375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,2048,0.29437867800394696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,2560,0.43759822845458984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,65536,14.463729858398438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,2048,0.155513776673211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,2560,0.4522017902798123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,1536,0.2350551022423638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,1024,0.1535724401473999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,1536,0.13140000237358943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,2048,0.49436177147759336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,2048,0.3710702260335286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,1536,0.2778186798095703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,768,0.11869511339399551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,1024,0.1106755534807841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,1536,0.29691465695699054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,1024,0.23209688398573133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,512,0.09081421958075629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,768,0.11353422535790338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,768,0.15780266125996908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,1024,0.22243287828233507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,256,0.0626879996723599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,768,0.1716782251993815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,512,0.09853511386447483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,512,0.12935022513071695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,512,0.1538791126675076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,256,0.08521688646740383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,256,0.11005866527557373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,64,0.04040888945261637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,256,0.09876977735095555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3584,128,0.08362044228447808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,128,0.10079022248586018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,32,0.042881776889165245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,128,0.07656711339950562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,64,0.09969955682754517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3584,32,0.10045244296391805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3584,128,0.05050933361053467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,65536,3.7354863484700522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,51200,7.456030951605903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,51200,3.0294098324245877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,65536,8.177662319607204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3584,65536,12.74285888671875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,16384,2.49290402730306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,16384,2.5411927964952254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,51200,8.797619289822048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,16384,2.6783493889702688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,16384,0.8054462009006076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,12288,0.6369226773579916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,12288,1.678711149427626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,51200,11.910586886935763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,10240,1.69250668419732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,12288,1.7563483979966905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,12288,1.9319591522216797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,65536,13.269913567437065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,10240,0.5078898006015354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,10240,1.5583502451578777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,8192,0.40983467631869847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,10240,1.6268577575683594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,8192,1.2789440155029297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,8192,1.2033857769436307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,7168,1.2106631596883137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,7168,0.3596213393741184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,8192,1.2644426557752821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,7168,1.0328533384535048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,6144,0.31662578052944607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,6144,0.9560079574584961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,6144,0.9102107153998481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,5120,0.6975022421942817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,7168,1.1244115829467773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,5120,0.32127200232611763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,6144,0.9846062130398221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,5120,0.7433724403381348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,4096,0.2480559878879123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,5120,0.7801164521111382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,3584,0.44162845611572266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,4096,0.6478755739000108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,3584,0.20558133390214708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,4096,0.9402977625528971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,3584,0.5274791187710232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,3072,0.4227155579460992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,3072,0.18162755171457926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,2560,0.3182151052686903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,3072,0.45104090372721356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,4096,0.5111920038859049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,3072,0.4772195816040039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,2560,0.18935910860697427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,2048,0.2506062189737956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,2560,0.47839821709526914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,2560,0.4039031134711371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,1536,0.2033182250128852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,2048,0.14999555216895208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,2048,0.30610132217407227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,1024,0.14110667175716823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,2048,0.33125776714748806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,1536,0.1379964484108819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,1536,0.24715733528137207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,768,0.10495733552508885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,3584,0.5528746710883247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,1024,0.17719733715057373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,1536,0.26531023449367946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,512,0.0842942198117574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,768,0.09706488582823013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,1024,0.20686311191982695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,768,0.14041866196526423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,256,0.05457511213090685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,768,0.15454310841030544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,512,0.11384088463253444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,512,0.08778488636016846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,128,0.03769244419203864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,256,0.08040622207853529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,256,0.09601333406236436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,256,0.08948622147242229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,128,0.0870435569021437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,128,0.07953333192401461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,32,0.03541244400872125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,128,0.06699822346369426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,64,0.0864586697684394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,3072,32,0.08647733264499241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,3072,1024,0.10109955734676784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,512,0.12259288628896077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,3072,64,0.033956444925732084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,65536,3.3154239654541016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,51200,2.6842693752712674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,51200,6.257613288031684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,65536,7.9615673489040795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,16384,1.7895378536648219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,16384,2.198404524061415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,16384,0.7515253490871854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,65536,9.72135755750868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,16384,2.323343065049913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,51200,7.680132548014323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,51200,9.656566196017796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,12288,0.5329199896918403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,10240,1.317267523871528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,12288,1.6151804394192164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,10240,1.2949022716946073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,12288,1.7131475872463648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,12288,1.450068473815918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,10240,0.541961775885688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,8192,0.9826089011298286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,10240,1.3914310667249892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,8192,0.35969244109259707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,3072,65536,10.916015625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,8192,1.0247484842936199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,7168,0.8694755766126844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,7168,0.3142942322625054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,8192,1.1192666159735787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,6144,0.6586382124159071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,6144,0.2752266724904378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,7168,0.9488560358683268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,5120,0.5590186648898655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,6144,0.765608893500434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,7168,1.1278799904717338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,5120,0.2436595492892795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,4096,0.48546579149034286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,5120,0.766224914126926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,6144,1.0961021847195096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,5120,0.6869520081414117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,4096,0.1993831131193373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,4096,0.5234062406751845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,4096,0.542901357014974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,3584,0.184242672390408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,3584,0.42467554410298664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,3072,0.33515556653340656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,3584,0.45487287309434676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,3072,0.16216177410549587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,2560,0.28226399421691895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,3072,0.3883368968963623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,3584,0.49113957087198895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,2560,0.1449271175596449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,2048,0.23775111304389107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,3072,0.4393048816257053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,2048,0.12706043985154894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,2560,0.3572942150963678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,65536,11.700691223144531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,2048,0.29509245024787056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,1536,0.21069955825805664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,1536,0.10807289017571343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,1024,0.115429335170322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,768,0.09001333183712429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,1536,0.23149421479966906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,1024,0.09255199962192112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,1024,0.15388533804151747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,512,0.06474577718310885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,768,0.0896071129375034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,768,0.12693333625793457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,768,0.1370968951119317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,2560,0.38576798968844944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,256,0.043939554029040866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,512,0.0809991094801161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,512,0.09788088666068183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,128,0.03634044528007507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,1536,0.17415644062889946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,512,0.10785600211885239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,2048,0.26695201132032603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,256,0.07339911328421699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,64,0.03187200095918443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,256,0.08190488815307617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2560,32,0.03356444504525926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2560,128,0.07133066654205322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,256,0.07704889112048678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,128,0.0743502246008979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,64,0.07344088951746623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2560,32,0.07272622320387098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,1024,0.16972177558475068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,65536,2.8020381927490234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,51200,3.911616007486979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,16384,1.5102755228678386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,51200,2.2645448048909507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,65536,6.290256924099392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,51200,6.399728139241536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,16384,0.6222595638699001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,16384,1.813808865017361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,12288,1.1405457390679252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,16384,1.9953592088487413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,65536,8.230322943793402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,12288,1.28895197974311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,12288,0.5976559850904677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,10240,0.9765911102294922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,51200,8.08425564236111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,12288,1.446896023220486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,10240,1.1078675587972004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,65536,10.093193901909723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,10240,0.5625288751390245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,10240,1.1787901984320748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,8192,0.7171013090345594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,8192,0.319102234310574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,7168,0.6028915511237251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,8192,0.8747377925448947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,7168,0.2701280117034912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,8192,0.9514533148871528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,7168,0.751979562971327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,6144,0.662176873948839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,6144,0.23909065458509657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,7168,0.8384586440192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,5120,0.42078044679429794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,6144,0.6537155575222439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,6144,0.7057369020250109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,5120,0.2061342265870836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,4096,0.3973084555731879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,5120,0.6139084498087565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,4096,0.17321156130896675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,4096,0.43312978744506836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,3584,0.3263537883758545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,3072,0.23722132047017416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,3584,0.15713511572943792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,4096,0.49285067452324766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,3584,0.38505866792466903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,3072,0.14125688870747885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,2560,0.211897783809238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,3584,0.4271128972371419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,3072,0.4413528972201877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,2560,0.13655377758873835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,3072,0.4012097782558865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,2560,0.2813653416103787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,2048,0.23147198888990614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,2560,0.3080480098724365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,1536,0.13542844189537898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,2048,0.11414755715264215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,5120,0.6051937739054362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,1024,0.08895555469724868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,2048,0.2625582218170166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,1536,0.17874488565656874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,1536,0.199837327003479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,1024,0.08491555849711101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,768,0.07598400115966797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,1024,0.13060711489783391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,1024,0.15005511707729763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,768,0.08167733086480035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,768,0.10336977905697292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,2048,0.17471199565463594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,768,0.11871822675069173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,512,0.0823777781592475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,256,0.03233244352870517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,512,0.0730480021900601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,128,0.029707554313871596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,512,0.0935644441180759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,256,0.06604711214701335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,256,0.06828177637524076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,64,0.024665777881940205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,256,0.06563378042644925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,128,0.0647875534163581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,32,0.025448888540267944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,2048,1536,0.09662044710583156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,128,0.06138222085105049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2048,128,0.050042665667004056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,64,0.0596382220586141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,2048,32,0.058858666155073375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,2048,512,0.05592711104287041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,65536,2.2827936808268228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,51200,3.608348422580295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,65536,4.308460235595703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,51200,1.7486168543497722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,16384,1.151426633199056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,16384,0.5543742179870605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,51200,5.25220701429579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,51200,5.714285532633464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,16384,1.4675181706746419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,12288,0.38656089040968156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,12288,0.8876853519015842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,16384,1.6492871178521051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,12288,1.0818790859646268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,65536,7.2120107014973955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,10240,0.6765724288092719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,12288,1.238740497165256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,10240,0.33048534393310547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,8192,0.5307768715752496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,10240,0.9158497916327583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,8192,0.25740710894266766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,10240,0.9983617994520398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,7168,0.539313793182373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,8192,0.9132347106933594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,7168,0.22496978441874185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,8192,0.7960426542494031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,7168,0.625399112701416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,6144,0.5426017973158095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,6144,0.2025697761111789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,7168,0.7147422366672092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,6144,0.5398275587293837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,6144,0.5954231156243218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,5120,0.1725191142823961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,5120,0.44164355595906574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,4096,0.2801297770606147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,5120,0.5099173121982151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,4096,0.14682932694753012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,3584,0.22156977653503418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,4096,0.35691555341084796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,5120,0.336916446685791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,3584,0.131059553888109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,3072,0.20346489217546251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,3584,0.32512354850769043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,3584,0.3925306797027588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,3072,0.12051999568939209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,3072,0.27088620927598744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,3072,0.314797321955363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,2560,0.23069689008924696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,2560,0.11124977800581191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,2048,0.13464177979363337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,2560,0.2666177749633789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,4096,0.4092115561167399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,1536,0.106513778368632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,2048,0.0985493328836229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,2048,0.18948533799913195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,2048,0.22088000509474012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,1024,0.07508889171812269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,1536,0.14677955044640437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,1536,0.08684533172183567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,2560,0.16924533579084608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,768,0.05638844437069363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,1536,0.17328443792131212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,1024,0.0750088890393575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,1024,0.10692977905273438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,512,0.04458311200141907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,768,0.07322133249706693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,1024,0.1274471150504218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,768,0.08483022451400757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,256,0.02676444417900509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,768,0.10072800185945298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,512,0.0653217766020033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,512,0.0669155584441291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,128,0.020954666866196528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,256,0.06020977762010363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,256,0.05412711037529839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,64,0.018325333793958027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,256,0.05483733283148872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,128,0.04852622085147434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,128,0.041448887851503156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,64,0.04614399870236715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1536,32,0.04500266578462389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,512,0.07846666706932916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1536,128,0.05875733163621691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1536,32,0.01889155473974016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,65536,3.1012933519151478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,65536,1.8515387641059027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,51200,2.512947506374783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,16384,0.7773937649197049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,51200,1.363184928894043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,65536,4.836019727918837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,16384,0.40680887964036727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,65536,5.429276360405816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,16384,1.1452657911512587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,51200,3.6105859544542103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,12288,0.605395581987169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,12288,0.3126595550113254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,10240,0.45525513754950625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,12288,0.8353004455566406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,12288,0.9704240163167318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,10240,0.2678648895687527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,8192,0.34762933519151473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,10240,0.697646247016059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,51200,4.355129665798612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,8192,0.20928177568647596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,16384,1.354207992553711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,7168,0.30878843201531303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,10240,0.8925448523627387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,8192,0.5668995645311143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,6144,0.26510132683648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,7168,0.2024071084128486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,8192,0.7361795637342664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,7168,0.49274577034844297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,6144,0.17557777298821342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,6144,0.41971466276380753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,5120,0.24519999821980795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,7168,0.5780115657382542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,4096,0.1865617831548055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,5120,0.13815555307600233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,6144,0.5845146709018284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,5120,0.34876444604661727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,4096,0.11866845024956597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,4096,0.28614311748080784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,3584,0.15606133143107095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,5120,0.4263937738206651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,3072,0.1465431054433187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,3584,0.10908533467186822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,3584,0.24545333120557997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,4096,0.4281368785434299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,2560,0.10619999965031941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,3072,0.21598754988776314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,3584,0.29869333902994794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,3072,0.29324621624416775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,2560,0.09318488836288452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,2560,0.1870702240202162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,2048,0.09300622012880112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,2560,0.22143377198113334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,1536,0.07264533307817247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,2048,0.08732355303234524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,2048,0.1477528942955865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,2048,0.1840346654256185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,1024,0.04943111207750109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,1536,0.07485422160890368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,1536,0.1150017711851332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1536,65536,6.873396555582683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,768,0.039319111241234675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,1536,0.14360088772243923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,1024,0.06674933433532715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,3072,0.10009777545928955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,1024,0.08378666639328003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,512,0.024832000335057575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,1024,0.10510755247539944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,768,0.06509688827726576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,768,0.06804000006781684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,256,0.018128888474570382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,768,0.08287822537951998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,512,0.05737600061628553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,128,0.014463999205165438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,512,0.06378133429421319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,256,0.0469359987311893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,256,0.04036710990799798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,64,0.012983111043771109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,256,0.043495999442206494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,1024,32,0.013588444226317935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,1024,128,0.0403511126836141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,128,0.034833778937657676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,1024,128,0.032637334532207914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,64,0.032567110326555036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,32,0.03120888935195075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,1024,512,0.05101244317160713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,51200,1.7195182376437719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,65536,1.7541697820027669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,65536,2.655591117011176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,16384,0.566739559173584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,51200,3.0302249060736766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,51200,1.2770577536688912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,16384,0.9886515935262045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,65536,3.916454315185547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,16384,0.39256978034973145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,12288,0.43195022477044004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,16384,1.204760021633572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,12288,0.30566665861341685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,51200,3.7775671217176647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,12288,0.7059653070237902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,65536,4.846699608696832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,8192,0.26372265815734863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,10240,0.5958782302008735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,12288,0.8838675816853842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,10240,0.2693102094862196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,7168,0.24680532349480522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,10240,0.7484729025099012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,10240,0.3511226707034641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,8192,0.4855208926730686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,7168,0.17928622828589547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,7168,0.4423697789510091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,6144,0.20999556117587617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,8192,0.6030969089931911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,6144,0.15663466188642713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,7168,0.5466497739156088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,6144,0.3679831027984619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,5120,0.3025004333919949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,4096,0.14294400480058458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,6144,0.45560089747111004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,5120,0.1353048880894979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,8192,0.2036355601416694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,3584,0.1257973379558987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,4096,0.1166035599178738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,5120,0.37980087598164874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,4096,0.24265599250793457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,3072,0.10983288950390285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,3584,0.1063457793659634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,4096,0.3146764437357585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,3584,0.21291466554005942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,2560,0.09000088771184285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,3584,0.27200267050001353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,3072,0.09847733047273423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,3072,0.18458310763041177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,2048,0.07527466615041097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,3072,0.24195999569363066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,2560,0.09214755561616685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,2560,0.15674400329589844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,1536,0.0549777779314253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,2560,0.20065422852834067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,2048,0.0841191079881456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,2048,0.12834044297536215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,1024,0.04016000032424927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,2048,0.16557955741882324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,1536,0.07385155227449205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,1536,0.09941689173380534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,768,0.032391998502943255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,1536,0.13016443782382542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,1024,0.06574577755398221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,1024,0.09457155730989243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,768,0.05660088857014974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,512,0.019290667441156175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,768,0.0640631119410197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,768,0.07382933298746745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,256,0.014187556174066333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,512,0.05335999859703911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,5120,0.18431111176808676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,512,0.04292177822854784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,128,0.011680000358157687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,512,0.056594669818878174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,256,0.0383840004603068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,256,0.03277066681120131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,64,0.010751999914646149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,256,0.038329776790406965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,128,0.028216888507207234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,768,32,0.011305777562989129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,768,128,0.027992000182469685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,64,0.02576711111598545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,32,0.024454222785101995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,768,1024,0.07161688804626465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,768,128,0.03621422251065572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,51200,1.3041573630438912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,65536,1.773190180460612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,65536,1.3824852837456598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,2560,128,0.05877600113550822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,16384,0.3603919876946344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,16384,0.31031645668877494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,51200,1.038185755411784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,12288,0.2612444559733073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,16384,0.9895938237508138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,65536,3.1888338724772134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,12288,0.24687822659810385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,10240,0.23197955555386016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,51200,2.5295147365993924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,12288,0.6207440164354112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,12288,0.7906079822116427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,10240,0.21296355459425184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,8192,0.18386133511861166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,10240,0.49533335367838544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,65536,4.192112816704644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,7168,0.1690151161617703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,8192,0.1601982249153985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,10240,0.668942239549425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,8192,0.39786399735344785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,16384,1.0368248621622722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,7168,0.14202933841281468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,8192,0.5226835674709743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,6144,0.12450488408406575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,5120,0.12398666805691189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,7168,0.4598506821526422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,6144,0.30098756154378253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,4096,0.0925368865331014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,6144,0.394686222076416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,5120,0.10938311285442776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,5120,0.25427288479275173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,3584,0.08319822284910414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,4096,0.09475111299090916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,5120,0.33132889535692006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,4096,0.20396978325313994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,3072,0.07055199808544584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,6144,0.14614666832817927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,4096,0.2693724367353651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,3584,0.08790400293138291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,3584,0.17996266153123644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,2560,0.06173066960440742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,7168,0.3573671181996663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,3072,0.08028266827265422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,3584,0.23793066872490776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,51200,3.29932000901964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,3072,0.15550400151146784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,2048,0.04946400059594048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,1536,0.03859822286499871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,3072,0.20756178432040742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,2560,0.13237867090437147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,2048,0.0695448915163676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,2560,0.17573066552480063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,1024,0.02605155607064565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,2048,0.10838399993048774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,1536,0.0639537771542867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,2048,0.1446346706814236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,768,0.020396444532606337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,1536,0.08368711339102851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,1024,0.05642311440573799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,1536,0.11400622791714138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,512,0.01443555619981554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,1024,0.06025155385335287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,768,0.05230399966239929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,1024,0.08180089129341973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,256,0.01054844425784217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,768,0.04689777890841166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,512,0.03493511014514499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,768,0.06392266352971394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,512,0.035432000954945884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,256,0.030273778571022883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,512,0.04787022206518385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,256,0.025733333494928148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,64,0.008240000241332585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,32,0.008399999803966945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,256,0.03173155585924784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,128,0.021732444564501446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,128,0.029824889368481104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,512,128,0.024296000599861145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,512,2560,0.07562488979763456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,64,0.01958755486541324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,512,32,0.018181333939234417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,51200,0.6415137714809841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,512,128,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,65536,0.9470560285780164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,16384,0.2677484353383382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,65536,1.3247111638387044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,51200,1.0058292812771268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,51200,2.0028809441460504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,16384,0.8113875389099121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,12288,0.16194221708509657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,65536,2.5751368204752603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,16384,0.3254480097028944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,10240,0.14327288998497859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,12288,0.2361831135219998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,16384,0.9260489145914713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,12288,0.48383378982543945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,8192,0.11321600278218587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,10240,0.2076408863067627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,10240,0.4321440060933431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,12288,0.6916791068183051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,7168,0.09761777851316665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,8192,0.32668712404039174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,10240,0.5745689074198405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,65536,3.64304690890842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,6144,0.09033333592944676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,7168,0.1368373367521498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,7168,0.28597866164313424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,8192,0.46199554867214626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,51200,2.8390818701850047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,5120,0.07057955529954699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,6144,0.12062754895952012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,7168,0.40644089380900067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,4096,0.05840533309512668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,6144,0.2472142113579644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,5120,0.10494666629367405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,5120,0.2070151170094808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,3584,0.051411555873023145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,6144,0.3489075501759847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,4096,0.09174399905734593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,5120,0.2922906610700819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,4096,0.16819822788238525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,3072,0.04423822296990288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,3584,0.08459466695785522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,3584,0.14731733004252115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,8192,0.1555751164754232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,2560,0.038415110773510404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,3072,0.07807377974192302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,3584,0.2098888953526815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,2048,0.027719111906157598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,3072,0.12751021650102404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,2560,0.07283911440107557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,3072,0.18222667111290825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,2560,0.10851111014684041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,2048,0.06703555583953857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,2560,0.15470400121476915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,2048,0.08883377578523423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,1024,0.013793778088357715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,2048,0.1277493370903863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,1536,0.0612657798661126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,1536,0.06793333424462213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,768,0.01149777736928728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,1536,0.09969688786400689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,1024,0.047578665945265025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,512,0.008319999608728621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,1024,0.07088977760738797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,768,0.03730666637420654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,768,0.0380311111609141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,256,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,768,0.054786665572060474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,512,0.02780710988574558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,512,0.031159109539455835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,4096,0.23772711224026152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,128,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,512,0.040722668170928955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,256,0.019278221660190158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,256,0.02571733295917511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,64,0.00583022211988767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,1536,0.02123111155298021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,128,0.029457777738571167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,128,0.015437333120240105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,64,0.01313688854376475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,256,32,0.012073777616024017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,1024,0.051098667913013034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,65536,0.6949457592434354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,256,256,0.029880000485314265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,51200,0.548416879442003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,256,32,0.0058266665372583605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,16384,0.17770489056905112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,65536,1.3284969329833984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,16384,0.5766311221652561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,65536,2.2807795206705728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,51200,1.0100800196329753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,12288,0.13411733839246961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,16384,0.3096133338080512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,51200,1.7831225925021703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,12288,0.46918757756551105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,10240,0.11527821752760146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,12288,0.2411875459882948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,10240,0.3680915567609999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,10240,0.21171200275421143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,8192,0.0920613341861301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,8192,0.2945653332604302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,8192,0.15872977839575872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,7168,0.08152621984481812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,7168,0.25782932175530326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,16384,0.8509475390116373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,7168,0.14028444555070665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,6144,0.07176799906624688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,12288,0.6427244610256618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,10240,0.5384524663289388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,6144,0.22239200274149576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,5120,0.06198044617970785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,6144,0.12060266070895725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,6144,0.3284471035003662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,5120,0.18637600209977892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,4096,0.04898222287495931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,5120,0.10489422082901001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,5120,0.2758435673183865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,65536,3.3647422790527344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,3584,0.04283022218280368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,4096,0.09138666921191746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,4096,0.15105777316623265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,4096,0.22470490137736002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,8192,0.43596622678968644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,3072,0.037830223639806114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,3584,0.08492177724838257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,51200,2.6280123392740884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,3584,0.1334302160474989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,2560,0.0328328874376085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,3584,0.19803910785251191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,3072,0.07850399944517347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,2048,0.027639110883076985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,3072,0.17215377754635283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,2560,0.09762933519151475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,2560,0.07247910896937053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,1536,0.01979733341270023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,2560,0.14589333534240723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,2048,0.0661440028084649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,2048,0.07970399989022149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,1024,0.009114666945404476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,2048,0.11957511636945937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,1536,0.06033155653211805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,1536,0.06056888898213705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,768,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,1536,0.09215644333097671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,1024,0.04327466752794054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,1024,0.0417431096235911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,512,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,1024,0.06337689028845893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,768,0.032454222440719604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,768,0.03420977791150411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,256,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,768,0.04907466636763679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,512,0.02367377777894338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,3072,0.11551288763682048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,7168,0.38152623176574707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,512,0.035636444886525474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,256,0.02983199887805515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,256,0.022982221510675218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,128,0.02955555584695604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,64,0.0045422220395671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,32,0.005043555465009478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,128,0.012110222544935016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,64,0.010019555687904358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,32,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,65536,0.6472258037990993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32768,128,512,0.031135112047195435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,51200,0.49380355411105686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,128,128,0.0042035554846127825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,128,256,0.015802666544914246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,16384,0.16279021898905435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,12288,0.12489867210388184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,10240,0.1027208897802565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,8192,0.08610933356814915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,12288,0.42407910029093426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,10240,0.35632001029120547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,16384,0.5786657863193089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,7168,0.073161780834198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,8192,0.2898142337799072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,6144,0.06286133660210504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,5120,0.05395999881956312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,4096,0.045180443260404796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,7168,0.25348088476392955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,6144,0.2179102235370212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,5120,0.18325866593254936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,3072,0.0328800015979343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,4096,0.1481297810872396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,2560,0.0265964451763365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,65536,2.2155787150065103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,51200,1.7296959559122722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,3584,0.13066932890150282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,1536,0.013854222165213691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,3072,0.11299467086791992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,1024,0.00811644477976693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,2560,0.09471111165152656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,2048,0.07603289021386041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,768,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,1536,0.057378669579823814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,512,0.005527110977305307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,1024,0.039447112215889826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,256,0.004243555582231945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,768,0.030778666337331135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,128,0.003526222374704149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,3584,0.03851822349760268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,256,0.01441422270403968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,64,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,32,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,128,0.010397333237859938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,64,0.009027555584907532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,32,0.009650666680600908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,64,2048,0.020735111501481798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,64,512,0.022231111923853557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,51200,0.5702942212422689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,16384,0.16643822193145752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,12288,0.12661155064900717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,10240,0.10556888580322266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,16384,0.563476456536187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,8192,0.08841600020726521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,12288,0.4238862196604411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,51200,1.723631117078993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,10240,0.3566471205817328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,6144,0.06549866994222005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,65536,0.7301200230916342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,5120,0.05527911252445645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,8192,0.2883004347483317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,65536,2.2188943227132163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,4096,0.04858666658401489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,7168,0.2533137798309326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,3584,0.04092266824510362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,6144,0.21813866827223036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,3072,0.03571288784344991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,2560,0.03039111031426324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,5120,0.18280177646213105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,2048,0.024313777685165405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,3584,0.1322480042775472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,3072,0.11464356051550971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,1536,0.014430221584108142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,1024,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,2048,0.07912621895472209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,768,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,7168,0.07637688848707411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,1536,0.06020622120963203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,1024,0.04183377822240194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,256,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,768,0.032758222685919866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,128,0.0028435554769304064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,512,0.023361777265866596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,64,0.0025031111306614345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,256,0.014482667048772177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,32,0.0042133331298828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,128,0.01038044442733129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,64,0.009951111343171861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,4096,0.1481191052330865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,32,0.008723555339707268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32768,32,2560,0.0962382223871019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32768,32,512,0.004549333204825719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,256,128,0.01997955474588606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32768,128,128,0.01868888901339637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,16384,8.592413160536024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,16384,24.92148166232639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,51200,26.56546698676215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,16384,29.641672770182293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,12288,17.474105834960938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,12288,6.441481696234809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,16384,42.456441243489586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,12288,22.617630004882812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,10240,13.521035088433159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,12288,30.639383951822918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,10240,5.419781155056423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,10240,18.19263373480903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,8192,10.450119018554688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,51200,65.85729302300346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,10240,25.589418199327255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,8192,4.1272400750054254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,7168,3.658471849229601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,7168,8.961204528808594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,8192,14.779837714301216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,8192,21.513980441623264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,6144,8.361718919542101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,7168,12.475498623318144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,6144,3.1447609795464406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,7168,17.564183553059895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,6144,11.052194383409288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,6144,10.61233435736762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,5120,6.877316368950738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,5120,2.5971306694878473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,5120,7.914054022894965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,4096,2.0206417507595487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,5120,7.624430338541667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,4096,5.214146508110894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,4096,6.038147396511501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,3584,1.8122213151719835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,4096,5.570286220974392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,3584,4.478784773084852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,3584,4.968603346082899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,3584,4.8317892286512585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,3072,4.211605495876737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,3072,3.9512354532877603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,3072,1.5982728534274633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,3072,4.160152859157987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,2560,1.3390898174709742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,2560,3.7581049601236978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,2560,3.4582676357693143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,2048,2.8323112063937717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,2560,3.4437963697645397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,2048,1.1705688900417752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,2048,2.8589856889512806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,2048,2.8609867095947266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,1536,0.8648764292399088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,1536,2.253679063585069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,1536,2.16792975531684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,1024,1.6432195239596898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,1024,0.7007804446750217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,1536,2.297090742323134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,768,1.0674435297648113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,1024,1.5848408804999456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,768,0.5845617718166775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,1024,1.586384031507704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,768,1.2445750766330295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,512,0.8314302232530383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,512,0.49244266086154515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,768,1.306505733066135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,512,0.9884942372639974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,256,0.527600023481581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,512,0.9913928773668078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,256,0.4759635395473904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,128,0.42638399865892196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,256,0.7255599763658312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,256,0.9714310963948568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,64,0.35945600933498806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,65536,128,0.4242222309112549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,128,0.8727093272738986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,65536,32,0.36403467920091415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,128,0.5571795569525825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,64,0.870549307929145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,32,0.9894382688734266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,65536,51200,127.81317816840277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,65536,26.676571316189236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,65536,51200,91.16917588975694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,51200,20.79988267686632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,51200,52.65582614474826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,16384,19.39148627387153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,16384,6.597277323404948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,65536,67.96662733289931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,12288,13.318685743543837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,16384,23.068037245008682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,12288,5.064687093098958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,16384,33.42737494574653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,12288,24.14428202311198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,12288,17.290035671657986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,10240,12.381279839409721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,10240,4.117867363823785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,51200,99.78238254123265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,10240,19.921512179904514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,8192,8.192058987087673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,8192,3.3502479129367404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,10240,14.629230075412325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,7168,7.068027496337891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,8192,11.555224100748697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,7168,2.7381119198269315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,8192,17.160166422526043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,7168,10.220141940646702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,6144,5.873445298936631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,6144,2.3738390604654946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,7168,13.834922790527344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,6144,9.819730970594618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,5120,4.929844326443143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,6144,8.039480421278212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,5120,2.04837777879503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,5120,6.557111104329427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,5120,6.02163569132487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,4096,4.09203126695421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,4096,1.6835813522338867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,4096,4.445468478732638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,4096,4.370288848876953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,3584,3.863524542914497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,3584,1.4079706403944228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,3584,3.905270046657986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,3584,3.797853257921007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,3072,1.1852871576944988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,3072,3.350878185696072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,3072,3.326418770684136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,3072,3.1938391791449656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,2560,1.031534194946289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,2560,2.9017111460367837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,2560,2.801230324639214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,2560,2.6797591315375437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,2048,2.384216096666124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,2048,0.8147199948628744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,2048,2.283568912082248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,2048,2.2676817576090493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,1536,1.7136702007717555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,1536,0.6563004387749566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,1536,1.7208986282348633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,1024,1.1873226165771484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,1536,1.7459199693467882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,65536,140.39922417534723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,1024,1.2600409189860027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,1024,0.6333049138387045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,768,0.896608034769694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,1024,1.188910272386339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,768,0.5050044589572483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,768,0.96578492058648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,512,0.6925813357035319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,768,0.9649724960327148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,512,0.7969226837158203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,512,0.36812623341878253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,256,0.5576213200887045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,512,0.7296897570292155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,256,0.337661345799764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,256,0.7098142306009928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,128,0.5033671061197916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,256,0.5590595669216579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,64,0.27368089887830943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,51200,32,0.28219376669989693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,51200,128,0.33315910233391655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,128,0.6614453527662489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,128,0.42791467242770725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,64,0.7326355510287814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,51200,32,0.772405359480116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,65536,91.97717963324652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,51200,51200,76.81456841362846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,51200,8.32633802625868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,65536,10.473558213975695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,16384,6.097379472520616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,51200,16.178780449761284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,16384,2.490658654106988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,65536,20.873402913411457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,16384,10.806233723958334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,12288,4.307214948866102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,16384,8.025157504611546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,12288,1.8721519046359594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,51200,23.346455891927082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,12288,5.784111870659722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,12288,7.80792236328125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,10240,4.162035624186198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,10240,1.501349343193902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,10240,4.6393962436252165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,65536,30.718366834852432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,8192,3.353343963623047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,8192,1.1016995112101238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,10240,6.488987392849392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,51200,32.4752197265625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,7168,0.8512541982862684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,8192,3.7017110188802085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,8192,5.535990397135417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,7168,2.489270316229926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,6144,2.195138719346788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,7168,3.1739243401421438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,6144,0.6997875637478299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,5120,0.6899395518832736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,6144,2.688162697686089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,7168,4.450626797146268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,5120,1.9302461412217882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,5120,1.9744311438666449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,4096,0.6128124660915798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,5120,1.963764402601454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,6144,2.8487023247612844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,4096,1.6321644253200953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,4096,1.4917404386732314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,4096,1.4235369364420574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,3584,0.4205999904208713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,3584,1.199194696214464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,3584,1.1765058305528429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,3072,0.36434043778313535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,3584,1.1434222327338326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,3072,1.2471884621514213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,2560,0.8856293360392252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,2560,0.32236356205410427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,3072,1.0206212997436523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,3072,1.0428213543362088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,2560,0.8311688635084363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,2048,0.27463022867838544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,2560,0.8000631332397461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,2048,0.8462950918409559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,1536,0.4982906447516547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,1536,0.22476445304022896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,2048,0.8894595040215386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,1024,0.4077591101328532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,1536,0.5286871062384711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,2048,0.6584951082865397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,1536,0.5231608814663357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,1024,0.16734577549828422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,768,0.32217333051893443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,1024,0.3811662197113037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,768,0.15290755695766872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,512,0.20735555224948457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,768,0.31408709949917263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,768,0.30962133407592773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,256,0.13901066780090332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,512,0.131951994366116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,128,0.10885244607925415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,256,0.14319022496541342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,512,0.4118231137593587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,512,0.25055021709865993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,256,0.22880265447828504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,256,0.191758222050137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,64,0.08834222290250991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,16384,128,0.12014044655693902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,16384,32,0.09230666690402561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,16384,128,0.15226133664449057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,1024,0.46912442313300234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,64,0.22339288393656412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,32,0.25190311008029515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,128,0.21826399697197807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,16384,65536,43.633073594835075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,65536,7.780853271484375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,51200,6.352853139241536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,16384,4.082575056287977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,51200,12.534605238172743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,16384,1.9319110446506078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,65536,15.122260199652779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,16384,8.174378289116754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,12288,3.2004434797498913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,16384,5.677088843451606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,12288,1.1435332828097873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,51200,18.201570298936634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,12288,5.833322736952041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,10240,2.710982216729058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,10240,0.9169644249810113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,12288,4.352613237169054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,8192,2.3169822692871094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,10240,3.456125259399414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,65536,23.470823499891495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,8192,0.6800595389472114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,10240,4.888970692952474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,8192,2.9652943081325955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,8192,2.9148311614990234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,7168,0.6049537658691406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,51200,24.729997422960068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,7168,1.8587396409776475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,6144,0.5083111127217611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,6144,1.9405439164903429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,7168,2.229556401570638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,7168,2.351515452067057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,6144,1.6905235714382594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,5120,0.432387563917372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,6144,2.0760089026557074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,5120,1.4786382251315648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,5120,1.3259252972073026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,4096,0.49228975507948136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,4096,1.0494826634724934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,4096,1.0143519507514107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,5120,1.4262773725721571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,3584,0.35892711745368105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,3584,1.007211579216851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,4096,1.069117334153917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,3584,0.8720151053534614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,3072,0.8252062267727323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,3584,0.883419566684299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,2560,0.6309946907891167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,3072,0.8307173517015246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,3072,0.7701653374565972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,2560,0.26425600051879883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,2560,0.6448239750332302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,2048,0.542527092827691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,2048,0.21065244409773085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,2048,0.5199928813510472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,2560,0.7697475751241049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,1536,0.3857582145267063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,3072,0.3654204474555121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,2048,0.5049582057529026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,1536,0.16557510693868002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,1024,0.27802756097581655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,1536,0.5725688934326172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,1536,0.39664533403184676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,1024,0.13062577777438694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,768,0.20551821920606825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,1024,0.3583235475752089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,1024,0.29063110881381565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,512,0.15748977661132812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,768,0.12341777483622234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,256,0.10338489214579265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,768,0.319102234310574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,512,0.10548266437318589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,512,0.20148889223734537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,768,0.24587999449835884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,128,0.08215733369191487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,512,0.1981297731399536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,256,0.09974222050772773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,256,0.17468533251020643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,64,0.07024444474114312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,256,0.15646399392022028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,12288,128,0.09536177582210964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,12288,32,0.07232444153891669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,128,0.16260355048709446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,12288,128,0.1114764478471544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,64,0.16384355227152506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,32,0.17614666620890299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,12288,65536,34.00576443142361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,65536,6.638181474473741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,51200,5.267071194118924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,16384,3.4404657151963978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,51200,12.997396681043837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,65536,13.743431091308594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,16384,6.779179467095269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,16384,1.3818719651963975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,16384,4.964070214165582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,51200,15.02390882703993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,12288,0.9801190694173177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,12288,3.0950516594780813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,10240,2.356863021850586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,10240,0.7843111356099447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,12288,3.5773883395724826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,12288,4.8727671305338545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,65536,19.95395745171441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,10240,3.0519457923041453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,51200,20.510815090603298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,10240,3.545064078436958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,8192,1.9234320322672527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,8192,0.6147564252217611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,8192,1.950432883368598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,7168,0.6243795818752712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,7168,1.6244391335381403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,6144,1.345378663804796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,7168,1.6767484876844618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,8192,2.56294674343533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,7168,2.1460933685302734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,6144,0.5050791104634603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,6144,1.25200441148546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,5120,1.0890817642211914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,5120,0.49033864339192706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,6144,1.7411929236518011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,5120,1.0691537857055664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,4096,0.966734250386556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,4096,0.3134391042921278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,5120,1.13590579562717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,4096,0.819801754421658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,3584,0.83096800910102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,3584,0.2711084418826633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,3584,0.7298124631245931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,4096,0.9085591634114584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,3072,0.6513288815816244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,3072,0.6243315802680122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,3072,0.24096179008483887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,3584,0.8217484686109754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,2560,0.2082648939556546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,2560,0.543632878197564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,2048,0.42148354318406844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,2560,0.563718212975396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,2048,0.21127377616034615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,1536,0.3512337737613254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,2560,0.5571786562601725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,2048,0.4625200165642633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,2048,0.41891911294725204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,1024,0.2084568871392144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,1536,0.18271999888949922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,1536,0.38845332463582355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,1024,0.2578870985243055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,768,0.18156533771091035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,1024,0.2555297745598687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,768,0.13795733451843262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,512,0.1261048846774631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,768,0.21671911080678305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,768,0.20374488830566406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,3072,0.6691813468933105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,256,0.08993333578109741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,512,0.09224977758195664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,512,0.22004622883266875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,512,0.16818666458129883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,128,0.0715164475970798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,256,0.08345955610275269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,256,0.14475556214650473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,1536,0.3503066698710124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,64,0.05769511063893636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,256,0.12474932935502793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,128,0.08460888597700332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,10240,32,0.060119999779595264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,10240,1024,0.11530844370524089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,128,0.13695732752482095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,10240,128,0.09812888834211562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,64,0.13771911462148032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,32,0.14534133010440403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,10240,65536,28.611226399739582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,51200,4.211297776963976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,65536,5.2340007358127165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,16384,2.737717310587565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,51200,8.548877292209202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,16384,0.9509288999769422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,65536,10.464726765950521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,16384,4.7953236897786455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,51200,10.146077473958334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,16384,3.1961502499050565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,12288,0.7159813245137533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,12288,2.074064042833116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,12288,2.476850721571181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,12288,3.084008958604601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,65536,12.966771443684896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,10240,2.0852542453342013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,10240,0.6134817865159776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,10240,1.9203626844618056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,10240,2.0745599534776473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,8192,0.47755024168226456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,8192,1.6656675338745117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,7168,1.245165295071072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,8192,1.4390426211886937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,8192,1.4200613233778212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,7168,1.229866663614909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,51200,15.595981174045138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,7168,0.4953013526068793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,7168,1.2549493577745225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,6144,0.8583013216654459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,6144,0.34830578168233234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,5120,0.2981315453847249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,6144,1.0356409284803603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,6144,1.0415208604600694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,5120,0.9872996012369791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,5120,0.8595342106289334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,4096,0.2918435467614068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,4096,0.6984666718377007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,5120,0.8476986885070801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,4096,0.6722844441731771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,3584,0.2199475500318739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,3584,0.6195466783311632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,4096,0.6632755597432455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,3072,0.47729868359035915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,3072,0.23520265685187447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,3584,0.6427217589484321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,2560,0.4233875539567735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,3072,0.5099031130472819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,3584,0.6085439787970649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,2560,0.17638222376505533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,3072,0.5054337713453505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,2560,0.473965326944987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,2048,0.4206399917602539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,2048,0.1476986673143175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,1536,0.26539378696017796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,2048,0.3567066722446018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,2560,0.5795626640319824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,1536,0.12048889531029595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,2048,0.3641653325822618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,1024,0.178619556956821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,1536,0.290720886654324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,768,0.14340089427100286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,1024,0.09551111194822524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,1024,0.20818311638302275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,512,0.09910666942596436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,768,0.09189244773652817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,1024,0.21521243784162733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,768,0.21038399802313912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,256,0.06714755296707153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,768,0.17117599646250406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,512,0.08006844255659315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,512,0.13448888725704616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,128,0.057130667898390025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,512,0.13787110646565756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,256,0.0718195570839776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,256,0.1236471070183648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,64,0.04628088739183214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,256,0.10588533348507351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,8192,128,0.07166310813691881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,8192,32,0.0476382209195031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,128,0.11044088999430339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,128,0.08200799756579928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,64,0.11276266309950088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,32,0.11560711595747207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,8192,1536,0.43958134121365017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,8192,65536,22.560991075303818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,65536,4.667919158935547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,51200,3.889996422661675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,16384,2.5730347103542752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,51200,8.399224175347223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,65536,9.415289137098524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,16384,0.8672791057162814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,51200,9.180196974012587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,16384,4.014632754855685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,16384,2.7928729587131076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,12288,1.9878968132866754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,12288,0.6025990910000271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,12288,2.3618337843153214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,65536,11.606155395507812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,12288,2.0052195654975042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,10240,1.7099919848971898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,10240,0.5023884243435329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,10240,1.7384462356567383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,8192,0.40320266617669004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,10240,1.6092898050944011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,8192,1.4294097688462999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,8192,1.2537404166327584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,8192,1.2977653079562717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,7168,0.45635732014973956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,7168,1.317090670267741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,7168,1.0902364518907335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,7168,1.0997982025146484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,6144,0.43439022699991864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,6144,0.9749999576144748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,51200,14.91505093044705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,6144,0.9290533065795898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,5120,0.7026586532592773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,6144,0.9018800523546008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,5120,0.3740791214836969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,5120,0.8024649090237088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,4096,0.2312213314904107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,4096,0.6125991079542372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,4096,0.5951911078559028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,3584,0.5624720255533854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,5120,0.832515557607015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,4096,0.634168889787462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,3584,0.2027804454167684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,3584,0.5498755772908529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,3584,0.5198817782931858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,3072,0.4601084391276042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,3072,0.18039111296335855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,3072,0.4604124493069119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,2560,0.3818897671169705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,2560,0.15768888261583117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,3072,0.45921601189507383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,2048,0.30764977137247723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,2560,0.39602043893602157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,2048,0.13271555635664198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,2560,0.500441763136122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,1536,0.221632891231113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,2048,0.36194578806559247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,1024,0.14957155121697321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,2048,0.3150373299916585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,1536,0.1076551079750061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,1536,0.25235022438897026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,768,0.1248942216237386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,1024,0.08600444263882107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,1024,0.19108711348639595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,1536,0.25061511993408203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,512,0.09128444724612766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,768,0.08033955759472318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,768,0.1459724505742391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,1024,0.20855466524759927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,256,0.06472000148561266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,768,0.15354755189683703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,512,0.07271022266811795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,512,0.11892088254292806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,128,0.04900177650981479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,512,0.12398489316304524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,256,0.10650933451122707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,256,0.0662942197587755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,64,0.04154755671819051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,256,0.09337155686484443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,128,0.09781689114040798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,7168,32,0.042973332934909396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,7168,128,0.065792891714308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,7168,128,0.07162400086720784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,64,0.09815644555621678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,32,0.10135644674301147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,7168,65536,19.276897854275173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,51200,3.251554701063368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,65536,3.8138249715169272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,16384,2.2033538818359375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,51200,6.764631059434678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,16384,3.0307892693413625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,16384,0.6934302118089465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,51200,7.8081554836697045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,65536,8.77344004313151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,12288,1.7921261257595484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,12288,0.5110746489630805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,16384,2.454620361328125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,12288,1.7971163855658636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,10240,0.440282662709554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,12288,1.7282124625311956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,65536,10.069813198513454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,10240,1.465568012661404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,10240,1.4367252985636394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,10240,1.4645137786865234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,8192,0.40251734521653915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,8192,1.2529564963446724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,8192,1.0430097579956055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,7168,1.029237323337131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,8192,1.1487412982516818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,7168,0.3765004475911458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,7168,0.9361119800143771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,7168,0.9515982733832465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,6144,0.7755146556430392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,6144,0.3248213397132026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,5120,0.644902229309082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,5120,0.2560737662845188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,6144,0.8326080110338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,51200,13.15641869439019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,6144,0.8136453098720975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,5120,0.6927084392971463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,4096,0.19986132780710855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,4096,0.6212639808654785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,5120,0.6616266568501791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,3584,0.4272631009419759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,4096,0.5692888895670573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,4096,0.523263136545817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,3584,0.22016533215840658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,3072,0.3565208911895752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,3584,0.4622151056925456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,3072,0.15744800037807888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,2560,0.30308710204230416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,3584,0.46245331234402126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,3072,0.5312817891438802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,2560,0.13794399632347956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,3072,0.3991217878129747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,2048,0.26841333177354604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,2560,0.3872613377041287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,1536,0.20738755332099068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,2048,0.11804266770680745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,2560,0.3387688795725505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,2048,0.28034043312072754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,1024,0.14355466100904676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,1536,0.09684266646703084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,2048,0.28163645002577037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,1536,0.31295021375020343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,1536,0.2246382236480713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,768,0.11206577883826362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,1024,0.16297422515021429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,1024,0.0789217750231425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,512,0.07875288857354058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,768,0.0753964450624254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,1024,0.1723768843544854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,768,0.18591022491455078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,256,0.05254044466548496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,768,0.1370497809516059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,512,0.10459199878904556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,65536,15.898963080512154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,512,0.06705777512656318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,128,0.03789955708715651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,512,0.11013955540127224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,256,0.06006222301059299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,64,0.03400533397992452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,6144,32,0.035368889570236206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,6144,128,0.05983289082845052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,256,0.08198222186830309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,128,0.09655021958880955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,6144,128,0.06339199675454034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,64,0.08475289079878066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,32,0.0870551135804918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,6144,256,0.11390844980875652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,65536,2.9799565209282766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,51200,2.3291439480251737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,16384,1.7765448888142903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,51200,5.457553863525391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,16384,0.5978888935512966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,16384,2.1466816796196833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,65536,7.439388699001736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,16384,2.0370915730794272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,12288,1.3752720091078017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,12288,0.4565484258863661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,51200,6.878441704644097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,12288,1.3437733120388453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,10240,0.3845662275950114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,65536,8.519983079698351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,12288,1.4515982733832464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,10240,1.1747830708821614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,8192,0.35748977131313747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,10240,1.165365325080024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,8192,0.9222284952799479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,10240,1.2212835947672527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,8192,0.8813581996493869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,8192,1.0001555548773873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,7168,0.27261421415540904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,7168,0.7724702093336316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,7168,0.8163910971747504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,6144,0.2819173336029053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,6144,0.7137982050577799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,7168,0.8495928976270887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,6144,0.6855609152052137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,51200,11.161403232150606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,5120,0.5299226442972819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,6144,0.6735902362399631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,5120,0.20314933194054496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,4096,0.4377022319369846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,4096,0.1714657809999254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,5120,0.6545999844868978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,3584,0.3659777906205919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,5120,0.6379226578606499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,4096,0.5529404216342503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,3584,0.15029066138797337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,4096,0.4917190869649251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,3072,0.32849333021375865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,3584,0.4598071310255263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,3072,0.13439288404252794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,3584,0.40069956249660915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,2560,0.26133065753512913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,3072,0.36136176851060653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,2560,0.11829777558644612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,2048,0.2204293277528551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,3072,0.3465235498216417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,2560,0.3695715533362494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,1536,0.1695635583665636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,2048,0.10199199782477485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,65536,13.246644761827257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,2048,0.2699671056535509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,1024,0.11317155096266006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,2048,0.24519732263353136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,1536,0.11038755708270603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,768,0.08867200215657552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,1024,0.0700133310423957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,1536,0.19385243786705864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,1024,0.1503173377778795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,512,0.06445511182149251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,768,0.06571022007200453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,768,0.1362204419242011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,256,0.04821688930193583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,768,0.12087910705142552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,512,0.05886755386988322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,512,0.08958666854434544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,512,0.09545066621568467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,128,0.03695022066434225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,256,0.05446488989724053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,256,0.07141600052515666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,2560,0.30189334021674263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,64,0.031854222218195595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,128,0.07176978058285184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,5120,128,0.053675555520587497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,1536,0.1869964467154609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,5120,128,0.05496444304784139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,5120,32,0.03365777929623922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,1024,0.14757333861456975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,64,0.07209155294630262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,32,0.07266933388180204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,5120,256,0.07737955782148573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,65536,2.4311627282036676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,51200,1.804019504123264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,16384,1.3385670979817708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,51200,4.444911956787109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,16384,0.5147635671827528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,65536,6.021415286593967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,16384,1.5788808398776586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,16384,1.6896791458129883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,12288,1.0708204905192058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,51200,5.4044587877061625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,12288,0.38981156879001194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,65536,6.95558844672309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,10240,0.9057475195990669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,12288,1.1576400332980685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,12288,1.366360028584798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,10240,0.3608977794647217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,10240,0.9768969217936198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,51200,8.019098917643229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,10240,0.9431813557942709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,8192,0.851463105943468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,8192,0.2612737814585368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,8192,0.744192017449273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,7168,0.6636782222323948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,8192,0.7201289070977105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,7168,0.29945866266886395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,6144,0.49872710969712997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,6144,0.19613422287835014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,7168,0.6768337885538737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,7168,0.6487173504299587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,6144,0.5481244193183051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,5120,0.434608883327908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,5120,0.17048711246914336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,4096,0.35346221923828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,5120,0.5629457897610134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,4096,0.14316977394951716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,6144,0.7913137541876899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,5120,0.4809608989291721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,4096,0.365739557478163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,3584,0.32048622767130536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,3072,0.2359022299448649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,3584,0.13833510875701904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,4096,0.5216257837083604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,3584,0.3398462136586507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,2560,0.2034311162100898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,3072,0.1304782231648763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,3072,0.2811644342210558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,3584,0.33736621008978945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,65536,11.018021477593315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,2048,0.1553271081712511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,2560,0.11260444588131374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,2560,0.2587066756354438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,3072,0.38008888562520343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,1536,0.130968001153734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,2048,0.19733422332339814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,1024,0.09118311272727118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,2048,0.10488622056113349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,1536,0.08229244417614408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,2048,0.20823200543721518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,768,0.07061422533459134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,1536,0.17209333843655059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,1024,0.07425689034991793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,1536,0.16754755708906385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,512,0.04985510971811083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,1024,0.11529066827562119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,768,0.058965331978268094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,1024,0.15028622415330675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,256,0.03422755665249295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,768,0.09230133559968735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,768,0.1015591091579861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,512,0.07462400197982788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,128,0.029847999413808186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,256,0.04833689000871447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,512,0.0834675563706292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,256,0.06936355431874593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,256,0.06155200137032402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,128,0.04761066701677111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,32,0.025123554799291823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,128,0.059168888462914355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,128,0.046160889996422656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,64,0.05894666910171509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,4096,2560,0.25231554773118764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,4096,32,0.05885599719153511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,4096,512,0.05256888932651944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,4096,64,0.02532088922129737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,65536,2.1843128204345703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,51200,4.076899634467231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,51200,1.643286175198025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,16384,1.347640037536621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,65536,5.79862297905816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,16384,0.5096435546875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,51200,4.672212388780382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,16384,1.3747146394517686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,16384,1.485855950249566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,12288,0.3390035629272461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,65536,6.251437293158637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,51200,6.540411631266276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,12288,1.0396515528361003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,12288,1.0502871407402885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,10240,0.7846755451626248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,10240,0.2919600009918213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,12288,1.0779120127360027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,10240,0.8488853242662219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,8192,0.22823643684387207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,8192,0.6804569032457141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,10240,0.8753493097093371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,8192,0.6600817574395074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,7168,0.22221955988142225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,8192,0.7236248652140299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,7168,0.5769422319200304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,6144,0.4716862042744954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,7168,0.6035920249091254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,6144,0.17744266986846924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,6144,0.49679819742838544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,5120,0.4011733267042372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,5120,0.1537839968999227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,6144,0.5200497839185927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,4096,0.3297768963707818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,5120,0.47624980078803164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,4096,0.13180622789594862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,7168,0.590686215294732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,65536,9.679345024956598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,5120,0.437101337644789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,3584,0.29529865582784015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,4096,0.362168894873725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,3584,0.11551200018988715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,4096,0.34738487667507595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,3072,0.26150933901468915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,3584,0.3120577865176731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,3072,0.10745688941743638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,2560,0.18641244040595162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,3072,0.25802577866448295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,3584,0.3141111003028022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,2048,0.17216000292036268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,2560,0.09425955348544651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,2560,0.22103110949198404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,1536,0.13102844026353624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,2048,0.0842026670773824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,2048,0.17824888229370117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,2560,0.2934808995988634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,1024,0.08489422003428142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,2048,0.19105777475568983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,1536,0.0741893317964342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,1024,0.05825510952207777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,768,0.06487555636299981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,1536,0.14972621864742702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,1024,0.10372533400853474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,512,0.04911199874348111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,1024,0.11931288242340088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,768,0.0547653337319692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,768,0.08592710892359416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,256,0.031303998496797346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,768,0.09229689174228245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,512,0.04946577880117628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,512,0.0661368899875217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,128,0.02663555575741662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,512,0.07397155629263984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,256,0.04495999879307217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,256,0.05679111348258125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,64,0.02199999988079071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,256,0.055008888244628906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,3072,0.2872062259250217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3584,128,0.04458489020665487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,128,0.052541332112418286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3584,128,0.0421066681543986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,64,0.05202133456865946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,32,0.05191733439763387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3584,1536,0.14047200149959987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3584,32,0.023181334137916565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,65536,1.8499031066894531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,51200,1.4594560199313695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,16384,1.1021617253621419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,51200,3.553229437934028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,65536,3.969610850016276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,16384,0.40541866090562606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,16384,1.2177688810560439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,12288,0.7625910970899793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,51200,4.1967243618435335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,16384,1.2288142310248482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,12288,0.3058942159016927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,12288,0.9063439899020724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,10240,0.6652408705817329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,65536,5.30492189195421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,12288,0.9369120068020291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,51200,5.7129626803927955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,10240,0.2609742217593723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,10240,0.7972764439053006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,8192,0.6875937779744467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,8192,0.20718488428327772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,10240,0.8613822195265027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,8192,0.6018595695495605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,7168,0.43881776597764754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,8192,0.6090364456176758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,7168,0.184608883327908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,6144,0.40262312359280056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,6144,0.16125067075093588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,7168,0.5301333533393012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,7168,0.7241075303819445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,6144,0.5717093149820963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,5120,0.35865423414442277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,6144,0.4676186773512099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,5120,0.13903021812438965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,4096,0.2627955542670356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,5120,0.37566932042439777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,4096,0.11690933174557155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,3584,0.20704800552792021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,5120,0.3840168846978082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,4096,0.3958551088968913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,3072,0.19051378303103975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,4096,0.31584445635477704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,65536,8.515872531467014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,3584,0.2627226776546902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,2560,0.18186399671766493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,3072,0.09633333153194851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,3584,0.2765057881673177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,3072,0.2830524444580078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,2048,0.13164622253841823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,3072,0.24943555725945366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,2048,0.07480622000164457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,2048,0.15943911340501574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,1536,0.0853759977552626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,2048,0.17119022210439047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,1536,0.06390044424268934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,1024,0.06882311238182916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,1536,0.12476978037092422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,3584,0.10760088761647542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,768,0.05377066797680325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,1024,0.05359911256366306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,1024,0.09201866388320923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,1024,0.10167111290825738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,512,0.04443822304407755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,768,0.0503768887784746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,768,0.07375466823577881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,2560,0.08585866954591538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,768,0.08204444249471028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,256,0.02688088847531213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,512,0.04564977685610453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,2560,0.2734942171308729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,512,0.06674666537178887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,128,0.020502222908867728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,256,0.042061332199308604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,256,0.04896355668703715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,64,0.018424888451894123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,3072,128,0.04160799913936191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,128,0.04619288775655958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,3072,32,0.019000889526473153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,1536,0.13563733630710179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,3072,128,0.03779733180999756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,64,0.04503555430306328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,32,0.04510666595564949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,2560,0.19183999962276885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,512,0.058436446719699435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,3072,256,0.04964533448219299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,65536,1.5962168375651042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,51200,1.2154755062527127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,16384,0.9725928836398654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,51200,3.0230577256944446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,65536,3.7565333048502603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,16384,0.38059645228915745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,16384,1.1948720084296334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,12288,0.7066791322496203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,16384,1.1369351281060112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,51200,3.6974461873372397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,12288,0.7788257598876953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,12288,0.2765724394056532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,12288,0.8343884680006238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,65536,4.71310297648112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,10240,0.5427884525722927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,51200,4.8165813022189665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,10240,0.2396942244635688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,8192,0.46568976508246523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,10240,0.6556231180826823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,8192,0.1840284400516086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,7168,0.3880586624145508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,65536,5.920632680257161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,8192,0.5306871202256944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,10240,0.678600894080268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,7168,0.1636684470706516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,6144,0.14226310782962376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,6144,0.3646728992462158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,5120,0.2565466562906901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,8192,0.6316480106777614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,7168,0.4718053076002333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,4096,0.2144977781507704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,5120,0.14536266856723362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,6144,0.4077102343241374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,5120,0.3255964385138618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,4096,0.10258489184909397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,4096,0.2577048937479655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,5120,0.41454044977823895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,3584,0.1896595557530721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,4096,0.30188711484273273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,3072,0.15288710594177246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,3584,0.09834488895204331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,3584,0.2271191014183892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,3584,0.24823111957973906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,3072,0.08904355764389038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,3072,0.19976533783806694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,7168,0.6234889030456543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,3072,0.23583910200330946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,6144,0.4357102182176378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,2048,0.10071911414464314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,2560,0.07778222031063504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,2560,0.16814933882819283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,1536,0.08444888724221124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,2048,0.06792800294028388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,2560,0.19576888614230684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,1024,0.05713599920272827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,2048,0.14267555872599283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,1536,0.05735911263359917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,2048,0.15389511320326063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,768,0.045477333996031016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,1024,0.04997155401441786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,1536,0.12156000402238633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,1024,0.07992977566189237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,512,0.032919999625947736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,1024,0.09124622080061172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,768,0.046462221278084644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,768,0.06426399946212769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,256,0.02313066687848833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,768,0.07397866911358304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,512,0.04218044545915392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,2560,0.1412986649407281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,128,0.017989334132936265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,512,0.059247109625074595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,256,0.0379688905345069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,256,0.042678223715888135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,64,0.01742844449149238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,256,0.04377511143684387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2560,128,0.03668444355328878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2560,128,0.03363555669784546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,64,0.03818044397566054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,32,0.038358221451441445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,1536,0.10853511095046997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,512,0.05096444487571716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2560,32,0.017948443690935772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2560,128,0.03941155473391215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,65536,1.27292357550727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,51200,0.9757209353976779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,16384,0.8755182160271539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,51200,2.4419697655571833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,65536,3.2476098802354603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,16384,0.30477245648701984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,16384,1.1715111202663846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,12288,0.5961573388841417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,16384,0.9453795750935873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,12288,0.6653084225124782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,12288,0.26249154408772785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,51200,3.3996675279405384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,65536,3.925176832411024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,10240,0.43265512254503036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,51200,3.140471988254123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,10240,0.199508441819085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,8192,0.3617573314242893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,12288,0.6814835336473254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,10240,0.5391973389519585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,65536,5.0845866733127165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,8192,0.15667022599114314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,7168,0.372303114997016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,10240,0.5917804506089952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,8192,0.49116086959838867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,7168,0.15631200207604304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,6144,0.2651662296719021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,8192,0.4840710957845052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,7168,0.38819021648830837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,6144,0.12270489003923203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,5120,0.21618755658467612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,7168,0.4079137908087836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,5120,0.10576266712612575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,4096,0.1905724472469754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,6144,0.5040159755282932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,6144,0.3596915668911404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,4096,0.089336891969045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,4096,0.2521599928538005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,3584,0.08193777667151557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,4096,0.24327021174960664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,3072,0.13196888234880236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,3584,0.19854399893018934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,3584,0.21743555863698324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,2560,0.10427466366026138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,3072,0.07824622260199653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,3072,0.168320894241333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,2560,0.07204355796178182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,2048,0.09461600250667995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,2560,0.14356799920399985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,2560,0.1643057796690199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,2048,0.12023288673824734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,2048,0.06041333410474989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,2048,0.13520356019337973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,3584,0.14967111746470133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,1536,0.07010222143597074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,5120,0.30610312355889213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,1536,0.05219466818703545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,1536,0.0928551091088189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,1536,0.10657511154810588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,1024,0.0681555536058214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,768,0.038809776306152344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,1024,0.045069333579805165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,1024,0.08004711071650188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,512,0.025013332565625507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,3072,0.18905866146087646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,768,0.04149511125352647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,768,0.055029332637786865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,256,0.017433777451515198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,768,0.06479999754163954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,512,0.035122666094038225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,512,0.04289866818322075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,128,0.014440889159838358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,5120,0.3010675642225477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,256,0.029263112280103896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,256,0.03587199913130866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,64,0.012925333446926542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,256,0.037646220790015325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,1024,0.04774577750100029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,2048,32,0.013556444810496436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,2048,128,0.028791109720865887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,128,0.03309511144955953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,128,0.028115553988350764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,64,0.03150755498144362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,2048,32,0.03121333320935567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,2048,512,0.05122488737106323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,65536,1.0778417587280273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,51200,1.9055663214789496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,51200,0.8403342035081652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,16384,0.6123058001200358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,65536,2.44921875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,16384,0.27598312166002065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,12288,0.46103911929660374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,51200,2.551022211710612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,16384,0.7167813513014051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,16384,0.8411715295579699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,65536,3.3898764716254344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,12288,0.5389742321438259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,10240,0.3370702266693115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,65536,3.3617305755615234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,10240,0.17263644271426729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,8192,0.27680712276034886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,12288,0.7136773533291287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,51200,2.7055316501193576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,10240,0.4607280095418294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,8192,0.14740710788302952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,7168,0.2320320076412625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,8192,0.36237155066596133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,10240,0.5155457920498318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,6144,0.1964319944381714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,7168,0.12087555726369222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,12288,0.20318755838606092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,7168,0.31892532772488064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,7168,0.3603057861328125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,5120,0.188263111644321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,6144,0.2742995686001248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,6144,0.10582400030559963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,6144,0.3109475506676568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,4096,0.13542221652136907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,5120,0.09115199910269843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,3584,0.12735111183590359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,5120,0.22567645708719888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,4096,0.07739644580417208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,4096,0.18073156144883898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,5120,0.2616159915924072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,3072,0.10253866513570149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,4096,0.2157448927561442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,3584,0.07196088631947835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,3584,0.16067111492156982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,2560,0.08347200022803412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,3072,0.0663742224375407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,3072,0.13924445046318903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,3584,0.20418132676018608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,2560,0.0609511137008667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,3072,0.16713778177897134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,2560,0.11934133370717366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,1536,0.05890133645799425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,2560,0.14152355988820395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,2048,0.05282310975922478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,8192,0.40803199344211155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,1024,0.03682400120629205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,2048,0.09710399972067939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,1536,0.04565599891874525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,1536,0.0768408907784356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,768,0.028832889265484277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,1024,0.0365973346763187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,1536,0.09223822090360854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,1024,0.05599911345375908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,512,0.019461333751678467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,1024,0.06883822547064887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,768,0.032758222685919866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,768,0.04566577739185757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,256,0.014408888088332282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,512,0.027290667096773785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,512,0.034824887911478676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,512,0.0444160004456838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,256,0.02612000041537815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,256,0.02880711025661892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,2048,0.06847288873460558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,256,0.032411555449167885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,64,0.01074044406414032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1536,128,0.02346933384736379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,32,0.011367999845080905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,128,0.024031110935741003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,2048,0.11904533704121907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,64,0.024698666400379602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,32,0.02437155610985226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1536,768,0.05561600128809611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1536,128,0.01181333346499337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1536,128,0.026344888740115698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,65536,0.8591670989990234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,65536,1.5476382573445637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,51200,1.276424937778049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,16384,0.40830310185750324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,51200,0.6882097986009387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,16384,0.20572711361779106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,12288,0.2968648804558648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,65536,2.5050977071126304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,16384,0.6986915800306531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,12288,0.16719288296169707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,65536,2.7205403645833335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,16384,0.7222293217976888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,10240,0.2235964404212104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,12288,0.4130391014946832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,8192,0.17785155773162842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,10240,0.1416071123547024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,12288,0.5190800031026205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,51200,2.1128658718532987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,10240,0.350400898191664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,8192,0.10866222116682266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,7168,0.15981688764360216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,8192,0.27768532435099286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,10240,0.4208853244781494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,6144,0.13095111317104763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,7168,0.09725599818759495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,8192,0.36126311620076496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,5120,0.12520533137851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,6144,0.09134755531946818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,51200,1.7734684414333766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,6144,0.21129688951704237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,7168,0.29492089483473033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,4096,0.10246222549014622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,6144,0.2558000087738037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,5120,0.17872799767388237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,3584,0.07842399676640828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,4096,0.06514844629499647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,4096,0.1438186698489719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,3584,0.05991199943754408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,4096,0.17568000157674155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,3584,0.12759999434153238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,3584,0.1569031079610189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,2560,0.056291553709242076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,3072,0.05585510863198174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,3072,0.11033511161804199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,3072,0.13736977842119005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,2048,0.047162665261162647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,2560,0.05154755711555481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,7168,0.2536995675828722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,2560,0.09415466917885675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,1536,0.03609777821434869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,2560,0.11619999673631455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,2048,0.045457777049806386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,5120,0.07704889112048678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,2048,0.07679288917117648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,1024,0.02418577836619483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,2048,0.09743111001120673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,1536,0.037808001041412354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,1536,0.060963557826148145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,768,0.019270223047998216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,5120,0.21765422821044922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,1536,0.07583110862308078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,1024,0.028318223026063707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,1024,0.044024000565210976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,768,0.02436888880199856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,1024,0.05570844146940443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,256,0.010425777898894416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,768,0.04524888926082187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,512,0.02215555475817786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,512,0.02741333345572154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,128,0.008857777549160851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,512,0.03510488735304938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,256,0.020879111356205408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,256,0.021976888179779053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,64,0.007895111209816402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,256,0.02641155487961239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,1024,128,0.020635555187861126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,32,0.008219555848174626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,128,0.019865777757432725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,1024,128,0.020277332928445604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,64,0.018260444204012554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,32,0.018048000004556444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,512,0.013471110827393003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,1024,768,0.035261332988739014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,65536,1.086893293592665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,65536,0.8269093301561145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,51200,0.9858311547173394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,1024,3072,0.0750675532552931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,16384,0.3483377827538385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,65536,1.9170390235053167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,51200,0.6604115698072646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,51200,1.4984479480319555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,16384,0.20197510719299316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,12288,0.1932551066080729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,16384,0.4759990904066298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,10240,0.1669315497080485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,65536,2.5182310740152993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,12288,0.16646488507588705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,16384,0.6385244263543023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,12288,0.3639715512593587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,8192,0.14098666773902044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,10240,0.14230399661593968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,12288,0.4600790871514215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,10240,0.30812356207105845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,51200,1.9272836049397786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,7168,0.1322737799750434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,8192,0.10473955339855617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,6144,0.10616711113188003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,8192,0.2398808797200521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,10240,0.387647098965115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,7168,0.09315555625491673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,5120,0.09556266996595596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,7168,0.21089511447482637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,6144,0.08198133442136976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,6144,0.1831564505894979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,4096,0.06724177466498481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,7168,0.2750053405761719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,5120,0.0720871090888977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,5120,0.1573866605758667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,3584,0.060990220970577665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,6144,0.27874310811360675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,4096,0.06308622492684259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,5120,0.2009377744462755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,3072,0.05223733186721802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,4096,0.12506400214301214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,3584,0.06068266762627495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,4096,0.16383111476898193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,2560,0.048001776138941445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,3584,0.11042222049501206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,3072,0.05434933304786682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,3584,0.1462951103846232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,2048,0.03894311189651489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,3072,0.0957546697722541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,3072,0.1275706688563029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,2560,0.08167733086480035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,1536,0.026340444882710774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,2048,0.04322310951020983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,2560,0.1109244426091512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,2048,0.06666933165656196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,1024,0.01899466746383243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,8192,0.3117537763383654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,2048,0.09023822016186184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,1536,0.03302133414480422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,1536,0.06975911060969035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,1024,0.025612443685531616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,1024,0.03809955716133118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,1024,0.05183822247717115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,768,0.02348355617788103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,768,0.04260266820589701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,512,0.023729777998394434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,256,0.00832622249921163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,512,0.02163999941613939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,512,0.0321742229991489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,2560,0.04983999994066027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,256,0.02057422200838725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,256,0.018725333942307364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,128,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,256,0.024217777782016333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,768,128,0.02030311193731096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,64,0.008405333591832055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,128,0.016513778103722464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,32,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,768,128,0.018567999203999836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,64,0.015140444040298462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,512,0.010844444235165914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,32,0.01482311056719886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,768,0.030361778206295435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,65536,0.6432275772094727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,51200,0.5725670920477973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,65536,0.6482844352722168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,768,768,0.014930667148696052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,16384,0.18551733758714464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,51200,0.5464195675320095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,51200,1.2433324389987521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,65536,1.6109689076741536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,16384,0.4009902212354872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,12288,0.13219200240241155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,16384,0.17309066984388563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,10240,0.11309066083696152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,12288,0.12780355082617864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,16384,0.553724447886149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,12288,0.3149973286522759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,768,1536,0.05266933308707344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,12288,0.4047608905368381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,51200,1.6672453350490992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,10240,0.11221244600084092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,10240,0.25657333268059623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,7168,0.08959022495481704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,8192,0.08452533351050483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,10240,0.33981066279941136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,8192,0.20387021700541177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,7168,0.07610755496554904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,6144,0.071834663550059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,8192,0.28254577848646373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,7168,0.17954933643341064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,5120,0.06496977806091309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,7168,0.24236000908745658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,6144,0.06749333275689019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,6144,0.15492978360917833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,5120,0.05936444467968411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,6144,0.20979377958509657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,5120,0.13042222128974065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,65536,2.153328789605035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,3584,0.046558221181233726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,8192,0.09349422322379218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,4096,0.05221955643759834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,4096,0.10674222310384114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,3072,0.0379342238108317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,3584,0.048828442891438804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,3584,0.0938746664259169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,2560,0.02972444560792711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,3584,0.12923910882737902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,3072,0.04523022307289971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,3072,0.08140710989634196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,2048,0.02437955637772878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,3072,0.11330310503641765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,2560,0.04114400015936957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,2560,0.06878399848937988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,1536,0.01936444474591149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,2048,0.03535022338231405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,2048,0.05635288688871595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,4096,0.048844443427191846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,1024,0.01367644468943278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,2048,0.0795102251900567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,1536,0.025302223033375208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,1536,0.04417066772778829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,768,0.011055111057228513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,5120,0.17772977881961396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,1536,0.06143199735217624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,1024,0.02144444485505422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,1024,0.03196000059445699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,512,0.008797333472304875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,1024,0.04500533474816216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,768,0.020257777637905545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,4096,0.14515911208258733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,256,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,768,0.02553244431813558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,768,0.03638222151332431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,512,0.018590221802393597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,128,0.006153777655627992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,512,0.020040000478426617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,256,0.017880888448821176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,512,0.027790221903059218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,256,0.01531733406914605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,32,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,256,0.020152888364262052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,512,128,0.017856889300876193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,128,0.013461333182122974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,128,0.01625333395269182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,64,0.012033777932325998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,512,2560,0.09597688913345337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,512,32,0.011752888560295105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,65536,0.44293777147928876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,51200,0.31761333677503795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,16384,0.09867199924257065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,65536,0.6268826590643989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,512,64,0.005865777946180767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,16384,0.35588799582587344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,51200,0.5365635553995768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,12288,0.08071821928024292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,51200,1.011522716946072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,65536,1.2860897911919487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,16384,0.15895022286309138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,10240,0.06988444593217638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,12288,0.1254133383433024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,12288,0.26238756709628636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,16384,0.47128799226548934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,8192,0.05628533495797051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,51200,1.4231928719414606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,10240,0.20763644907209608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,12288,0.34800444708930117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,65536,1.8479359944661458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,8192,0.08260533544752333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,10240,0.29294310675727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,8192,0.1693413390053643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,6144,0.04273244407441881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,8192,0.24072533183627656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,7168,0.14800355169508192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,5120,0.035819556978013777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,6144,0.12777066230773926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,6144,0.06589155726962619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,7168,0.21277244885762533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,6144,0.18114399909973145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,4096,0.029735998974906072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,5120,0.10799288749694824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,5120,0.05792355537414551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,5120,0.1537137826283773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,4096,0.050684445434146456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,3584,0.025951999757024977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,10240,0.11082666450076634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,4096,0.08781955639521281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,7168,0.0478097763326433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,3072,0.02355555527740055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,4096,0.125072889857822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,3584,0.04741866721047295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,2560,0.017208000024159748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,3584,0.07723289065890841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,3072,0.04398222102059258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,3584,0.11168977949354385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,3072,0.06688621971342298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,2048,0.01386755539311303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,7168,0.07498221927218966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,3072,0.09812711344824897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,2560,0.03985777828428481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,1536,0.011063999599880643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,2560,0.05632177988688151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,2048,0.030185778935750324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,2560,0.08365511231952244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,1024,0.008714666797055138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,1536,0.023595554961098567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,2048,0.06880088647206624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,1536,0.03632355398601956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,768,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,1024,0.021009777983029682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,512,0.006031999985376994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,1024,0.02576444380813175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,1024,0.03733955489264594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,768,0.01992177797688378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,768,0.020972443951500788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,768,0.030893332428402368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,512,0.018552889426549275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,512,0.016241777274343703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,128,0.00388711111413108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,512,0.024001777172088623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,256,0.01778133379088508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,256,0.012061333490742577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,64,0.003549333247873518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,256,0.01759822169939677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,128,0.010010666317409938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,32,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,128,0.013456888496875763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,64,0.008744888835483128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,32,0.008625778059164682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,256,2048,0.04603999853134155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,65536,0.33082932896084255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,256,1536,0.0525911119249132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,51200,0.26141778628031415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,256,256,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,16384,0.0839395523071289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,65536,0.6342622439066569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,256,128,0.017525333497259352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,51200,0.5109893480936686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,16384,0.2985360092586941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,65536,1.1626301871405709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,12288,0.06797599792480469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,51200,0.941581302218967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,16384,0.15903288788265652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,10240,0.05468711256980896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,12288,0.1288248830371433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,12288,0.22544889979892305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,16384,0.43800444073147243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,8192,0.046315554115507335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,10240,0.11390311188167995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,65536,1.6850222481621637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,51200,1.3194284439086914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,12288,0.3278977870941162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,7168,0.039537777503331505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,10240,0.2759671211242676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,6144,0.03417777683999803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,8192,0.085372441344791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,8192,0.15456977155473497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,7168,0.07482844591140747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,5120,0.02976800004641215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,7168,0.135660449663798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,6144,0.06543733014000787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,7168,0.19633777936299643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,6144,0.11746577421824138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,4096,0.024624889095624287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,5120,0.05795733133951823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,5120,0.09897244638866848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,3584,0.02160355614291297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,4096,0.05059911145104302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,5120,0.1431164476606581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,4096,0.07980444696214464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,3072,0.018375110295083787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,4096,0.11630577511257595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,3584,0.04748533169428507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,10240,0.18956266509162056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,3584,0.07022400034798516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,2560,0.01017599966790941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,3584,0.10367377599080403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,3072,0.04336977667278714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,3072,0.09064444568422104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,2560,0.05099822084108988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,2560,0.03927910990185208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,8192,0.2252826690673828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,2560,0.07695377535290189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,1536,0.007170666423108842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,2048,0.028972443607118394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,2048,0.041386667225095965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,2048,0.06241955359776815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,1536,0.03211822112401327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,1536,0.023771555887328252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,1536,0.04804355568355984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,1024,0.02306933369901445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,768,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,1024,0.033927109506395124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,3072,0.06100355254279243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,768,0.018645332919226754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,768,0.02013066742155287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,2048,0.008446222378147973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,512,0.0042151109211974675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,768,0.028139554791980322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,256,0.00349866681628757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,512,0.018611555298169453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,512,0.014663111832406787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,512,0.021643555826610986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,256,0.01038933296998342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,128,0.002883555574549569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,256,0.017893332574102614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,1024,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,64,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,256,0.0158924443854226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,128,32,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,128,0.017551110850440133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,128,0.008290666672918532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,64,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16384,128,1024,0.021144888467258875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,128,32,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,51200,0.2483724488152398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,65536,0.3225822183820936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,12288,0.06492088900672065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,10240,0.05284355415238274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,16384,0.28768356641133624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,12288,0.22369334432813856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,8192,0.04603022336959839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,7168,0.03820266657405429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,10240,0.18252178033192953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,6144,0.033430222007963396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,8192,0.1507048871782091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,51200,0.8690284623040093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,65536,1.120774269104004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,7168,0.13030489285786948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,4096,0.023991111252042983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,3584,0.02090222140153249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,16384,0.08127999967998929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,5120,0.09434488746854995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,3072,0.016068443655967712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,4096,0.07589955462349786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,3584,0.06656799713770549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,2048,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,3072,0.05697866943147448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,2560,0.048136889934539795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,1536,0.00758755538198683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,1024,0.005771555420425203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,2048,0.03933599922392104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,768,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,1536,0.0303457776705424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,512,0.004214222232500712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,768,0.01788444485929277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,5120,0.028439998626708984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,256,0.003528888854715559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,128,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,6144,0.11255289448632134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,64,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,512,0.013818666338920593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,32,0.0026177778426143858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,256,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,128,0.00742133292886946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,64,2560,0.00963644435008367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,64,0.006565333240562015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,32,0.007071110937330458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,16384,0.07921866575876872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,65536,0.3561600049336751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,51200,0.2563146750132243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,12288,0.05984977881113688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,10240,0.05065600077311198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,64,1024,0.02199999988079071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,16384,0.28721777598063153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,8192,0.041259557008743286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,7168,0.03636444608370463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,12288,0.2168204387029012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,6144,0.03211466802491082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,10240,0.1826444466908773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,51200,0.8691928651597766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,5120,0.027416000763575237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,8192,0.14863022168477377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,4096,0.022541332576009963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,3584,0.019555555449591745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,7168,0.1299475563897027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,3072,0.014694222145610385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,6144,0.11230666107601589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,5120,0.09415377510918511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,4096,0.0758568909433153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,2560,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,2048,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,3584,0.07057866785261366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,1536,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,3072,0.060425778230031334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,1024,0.004287999951177173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,2560,0.050833778248892895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,2048,0.0414373344845242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,768,0.003879111260175705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,512,0.0032231110251612137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,1536,0.032257778777016535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,256,0.002586666701568498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,1024,0.022872888379626807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,128,0.00245066670080026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,768,0.018343110879262287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,512,0.014121777481502958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,64,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16384,32,32,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,256,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,65536,1.1202657487657335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,128,0.0074684446056683855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,64,0.006646222124497096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16384,32,32,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,16384,10.226002163357204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,51200,13.34161122639974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,128,0.012732444538010491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16384,128,6144,0.1701528893576728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,16384,4.04311286078559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,16384,18.12232123480903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,12288,2.8443351321750217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,12288,7.8230031331380205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,16384,13.348108079698351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,12288,11.933019002278646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,12288,9.558156331380209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,10240,7.2254223293728295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,10240,2.348111894395616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,51200,32.55390421549479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,10240,8.20054202609592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,10240,7.463375515407986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,8192,1.78294923570421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,8192,5.856250762939453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,8192,5.322743309868707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,7168,1.5877564748128254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,7168,4.6481272379557295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,8192,5.5196745130750875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,7168,4.50397703382704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,6144,3.884677463107639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,7168,4.540733337402344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,6144,1.3815217547946508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,5120,1.1542960272894967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,5120,3.2001991271972656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,6144,4.200659434000651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,6144,3.9827838473849826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,5120,3.1768853929307728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,4096,0.927238252427843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,5120,3.277147505018446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,4096,2.757220374213325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,4096,2.6676701439751516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,4096,2.693093405829536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,3584,2.230560090806749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,3584,0.8018515374925402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,3584,2.4215111202663846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,3072,1.893904791937934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,3072,0.7294409010145398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,3584,2.341471142239041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,3072,1.976943122016059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,2560,0.6289128727383083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,3072,2.039366192287869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,2560,1.7778684828016493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,2048,1.3047635820176866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,2560,1.6095342636108398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,2560,1.690102259318034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,2048,0.543368869357639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,2048,1.3621484968397353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,1536,0.9584692849053277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,2048,1.3759164810180664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,1536,0.40779466099209255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,1536,1.0412488513522677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,1024,0.6717768775092231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,1024,0.3612817658318414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,1536,1.06749513414171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,768,0.47800265418158633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,1024,0.7992204560173882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,51200,45.31138780381945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,768,0.27229155434502494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,1024,0.7250906626383463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,512,0.3873804410298665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,768,0.6879182391696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,256,0.26793334219190806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,768,0.7116266886393229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,512,0.2381582260131836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,512,0.5093386438157823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,256,0.21759466330210367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,128,0.2377164363861084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,256,0.371321784125434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,256,0.468864017062717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,512,0.460935115814209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,64,0.20197955767313638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,128,0.21746933460235596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,128,0.2751164436340332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,65536,32,0.19115822845035127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,64,0.44968356026543516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,32,0.4692693286471897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,128,0.4389493465423584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,51200,63.45460340711805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,65536,13.272904290093315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,51200,10.166883680555555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,16384,9.827210320366754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,51200,25.899122450086807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,16384,3.0455137888590493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,65536,33.06310356987847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,16384,10.342774285210504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,12288,6.317546844482422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,16384,14.357313368055555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,12288,2.198380364312066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,51200,35.692657470703125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,12288,9.111753675672743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,10240,5.490427652994792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,12288,7.304400126139323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,10240,1.7148959901597765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,10240,6.553469340006511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,8192,4.10859637790256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,8192,1.5313715404934352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,10240,5.672332339816624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,8192,4.277743869357638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,8192,4.387140485975477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,7168,1.287674691942003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,7168,4.05204815334744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,51200,49.686309814453125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,7168,3.881986829969618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,6144,0.9744319915771484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,65536,46.988427056206596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,7168,3.6179305182562937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,6144,3.3706658681233725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,6144,3.1621795230441623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,5120,0.8808088832431369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,5120,2.5575129191080728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,6144,2.979280047946506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,4096,1.9308079613579645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,5120,2.6421705881754556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,4096,0.7198293473985461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,4096,2.0933119455973306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,5120,2.4870391421847873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,3584,0.6824924680921766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,4096,1.9917892879909938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,3584,1.9462240007188585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,3584,1.8386249542236328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,3072,0.5227920214335123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,3584,1.8258018493652344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,3072,1.6401857799953883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,3072,1.5089555316501195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,2560,1.4121600257025824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,2560,0.4711626900566949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,3072,1.5332258012559679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,2560,1.2698515786064994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,2048,1.0616221957736545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,2048,0.43569689326816136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,2560,1.2891324361165364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,2048,1.0212862226698134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,1536,0.32455465528700084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,1536,0.8047964307996961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,1024,0.5193849139743382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,2048,0.9700408511691623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,1536,0.8028177685207791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,1024,0.6037706798977321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,768,0.3972240024142795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,1024,0.5646213425530328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,768,0.2148240009943644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,768,0.507056024339464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,512,0.35764800177680117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,768,0.45734845267401797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,512,0.18492178122202554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,256,0.3088000085618761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,512,0.4033786720699734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,512,0.36452444394429523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,256,0.3682435618506537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,1536,0.8610542085435656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,128,0.168250666724311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,256,0.26867111523946124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,256,0.2946257856157091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,128,0.1666204399532742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,64,0.15007732974158391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,51200,32,0.14465155866410997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,128,0.3408408959706624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,51200,128,0.2213519944085015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,64,0.3571733368767633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,32,0.36711645126342773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,51200,1024,0.3023644553290473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,65536,5.190503862169054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,51200,65536,69.66014946831596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,51200,8.03823513454861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,65536,10.170353359646267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,51200,4.075362735324435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,16384,3.1263946957058377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,16384,1.170471085442437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,12288,2.02364624871148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,16384,4.6812693277994795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,51200,11.894913567437065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,16384,3.2604736751980252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,12288,2.9188444349500866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,12288,0.860982206132677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,12288,2.3416194915771484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,10240,1.7269137700398762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,10240,0.6530826886494955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,65536,15.030349731445312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,8192,1.5675217310587566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,51200,16.165430704752605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,10240,2.054335912068685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,10240,1.9580213758680556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,8192,1.4138782289293077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,8192,1.3848071628146703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,8192,0.5297653410169814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,7168,0.4249386787414551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,7168,1.3031031290690105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,7168,1.2193102306789823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,6144,0.9261084662543403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,7168,1.22179200914171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,6144,0.38887911372714573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,6144,0.9912622239854602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,6144,0.9697831471761068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,5120,0.30695732434590656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,5120,0.9009306165907117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,4096,0.6852240032619901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,5120,0.821033795674642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,4096,0.23708711730109322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,5120,0.7791662216186523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,4096,0.6453973452250162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,3584,0.21353422270880806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,4096,0.6364515622456869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,3584,0.6661973529391819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,3072,0.5063261985778809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,3584,0.5381360054016113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,3072,0.18500710858239067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,3584,0.6650542153252496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,3072,0.49004533555772567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,2560,0.16101866298251682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,2560,0.4690275722079807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,3072,0.47112533781263566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,2048,0.3614853223164876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,2048,0.17210755083296034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,1536,0.2771857844458686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,2560,0.4294115437401666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,2560,0.6527893278333876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,2048,0.34246667226155597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,1024,0.17794488535987008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,1536,0.14646489090389675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,1536,0.43661244710286456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,768,0.1544204420513577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,1024,0.10070044464535183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,1024,0.3173031012217204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,1024,0.19348355134328207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,768,0.09211911095513238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,512,0.10510310861799453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,768,0.27383822864956325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,256,0.07553511195712619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,512,0.13402667310502794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,768,0.24742311901516387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,65536,21.410651312934025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,512,0.133151994811164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,128,0.06369866927464803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,256,0.09345066547393799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,256,0.1203111145231459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,256,0.10046399964226617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,2048,0.32751109864976674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,64,0.046281778150134616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,128,0.0956160028775533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,16384,32,0.047911110851499766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,128,0.11448444260491265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,128,0.08698755502700806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,1536,0.29984354972839355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,64,0.11682132879892985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,32,0.12196355395846897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,512,0.07009066475762261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,51200,2.9654738108317056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,65536,3.814451429578993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,16384,2.3100221421983504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,51200,6.3125805324978295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,16384,0.8592400021023221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,16384,3.5068160163031687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,16384,2.474755605061849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,65536,8.190267774793837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,12288,1.7766319910685222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,51200,9.081012301974827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,12288,0.5465857717725966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,12288,2.145873811509874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,10240,0.4317333433363173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,10240,1.2595396041870117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,12288,1.8593031565348308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,10240,1.4977484809027777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,10240,1.4404658211602106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,8192,1.1573173734876845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,8192,0.34749865531921387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,65536,11.442782931857638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,8192,1.0917297999064128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,51200,12.025677998860678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,7168,0.9047732883029513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,7168,0.43787821133931476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,8192,1.0721413294474285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,6144,0.8235893249511719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,7168,1.011305809020996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,7168,0.9164915084838867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,6144,0.2641493214501275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,5120,0.2204808923933241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,5120,0.759426646762424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,6144,0.7669653362698026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,5120,0.6399733225504557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,4096,0.5071839756435818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,5120,0.6176097657945421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,4096,0.18191288577185738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,6144,0.7597733073764377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,4096,0.4800657696194119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,3584,0.4761217965020074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,4096,0.5126035478379991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,3584,0.16033066643608943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,3584,0.4475626415676541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,3584,0.4093715614742703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,3072,0.3621315426296658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,3072,0.141957336001926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,2560,0.30401600731743705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,2560,0.12383200062645806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,3072,0.5626693301730686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,3072,0.35960621303982204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,2560,0.3223866621653239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,2048,0.2627351019117567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,1536,0.19737599955664742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,2560,0.3360648949940999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,2048,0.1399093336529202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,2048,0.29286132918463814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,1024,0.1371199952231513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,1536,0.08789155880610149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,1536,0.20643111069997153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,2048,0.25103643205430776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,1536,0.20068711704678008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,768,0.1064515577422248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,1024,0.07193066676457723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,1024,0.15753422843085393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,512,0.07754755682415433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,1024,0.15366578102111816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,768,0.06143999762005276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,256,0.05606310897403293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,768,0.12325867017110188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,512,0.0638293359014723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,768,0.12627111540900335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,128,0.044433779186672635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,512,0.10566666391160752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,256,0.05112533436881172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,512,0.10247021913528442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,256,0.09293866819805569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,64,0.03546311126814948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,12288,32,0.03567644291453891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,256,0.07894577582677205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,128,0.0517520010471344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,128,0.08679999907811482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,128,0.06227822436226738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,65536,17.119613647460938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,32,0.09322577714920044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,64,0.08930133448706733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,65536,3.1383910708957248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,51200,2.580054177178277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,16384,1.8655918969048395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,51200,5.622417873806423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,16384,0.6739110946655273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,65536,7.011329650878906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,16384,2.9613414340549045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,51200,7.136600918240017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,16384,2.0175537533230252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,12288,1.4353698094685872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,12288,0.47583733664618594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,12288,1.8177074856228297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,10240,1.1015289094712999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,12288,1.4561857647365992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,10240,0.3548080126444499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,10240,1.2887724770439994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,65536,9.445621066623264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,8192,0.2852524386511909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,10240,1.19868893093533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,8192,0.958006223042806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,8192,0.8553991317749023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,51200,10.153499179416233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,7168,0.2696888976626926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,8192,1.1315093570285373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,7168,0.8563644621107313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,6144,0.2199315494961209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,7168,0.7860870891147189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,6144,0.6628773477342393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,7168,0.8950044843885633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,6144,0.6497519810994467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,6144,0.6277937889099121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,5120,0.21327555179595947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,5120,0.5742399957444932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,4096,0.41352444224887425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,5120,0.5621822145250108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,5120,0.492686218685574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,4096,0.1781057781643338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,3584,0.4038835631476508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,4096,0.4228773381974962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,3584,0.13981066809760198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,3072,0.2997937732272678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,4096,0.40517600377400714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,3584,0.400802665286594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,3584,0.3504862255520291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,3072,0.13872799608442518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,2560,0.2497946686214871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,3072,0.3996933301289876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,3072,0.30359289381239146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,2560,0.1117066674762302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,2048,0.20778577857547334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,2560,0.29583289888170033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,1536,0.16264266437954372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,2048,0.09172266721725464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,2048,0.23075999153984916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,2048,0.21573332945505777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,1536,0.07372355461120605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,1536,0.17605155044131807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,1024,0.11974844667646621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,768,0.0928293334113227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,1536,0.19975822501712373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,1024,0.059502224127451576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,1024,0.13197510772281223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,1024,0.13146400451660156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,512,0.06763911247253418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,768,0.11130221684773763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,65536,13.849582248263888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,768,0.11173866854773627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,256,0.04576177729500664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,512,0.049703111251195274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,512,0.08812889125612046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,128,0.036899556716283165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,512,0.08831555313534206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,2560,0.25911198721991646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,256,0.04577066832118564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,64,0.03119022316402859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,256,0.07163555754555596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,10240,32,0.0340008901225196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,128,0.04493688874774509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,128,0.07361866368187799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,128,0.0519644452465905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,64,0.07562400235070123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,32,0.07916000154283312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,768,0.05396799908743965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,256,0.07848978042602539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,65536,2.6096969180636935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,51200,1.9346302880181208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,16384,1.445677333407932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,51200,4.463383144802518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,65536,4.944754706488715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,16384,0.5253164503309462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,16384,2.4385253058539496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,12288,1.0350799560546875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,51200,5.654634687635634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,12288,0.35182666778564453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,16384,1.6749191284179688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,12288,1.5135537253485785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,12288,1.1501493453979492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,10240,1.0610550774468315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,10240,0.2889333301120334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,10240,1.0500497817993164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,8192,0.7402710914611816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,65536,7.530426025390625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,8192,0.23079199261135527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,10240,1.0172320471869576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,51200,8.194293551974827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,8192,0.7329146597120496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,8192,0.7934702237447103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,7168,0.6609635353088379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,6144,0.45938846800062394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,6144,0.21531466643015543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,7168,0.687214215596517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,6144,0.4831386672125922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,6144,0.6196791330973307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,5120,0.15373422039879694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,7168,0.22494488292270234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,4096,0.3342906634012858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,5120,0.6173315578036839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,4096,0.12868711683485243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,3584,0.2862791220347087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,7168,0.5897448857625326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,4096,0.3527395460340712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,4096,0.3274471229977078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,3584,0.1164355542924669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,3072,0.24999732441372344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,3584,0.311901330947876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,5120,0.42524533801608616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,3584,0.3025217851003011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,3072,0.10119377904468113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,2560,0.22252088122897676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,3072,0.25987111197577584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,2560,0.08862844440672134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,2048,0.17267466915978325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,5120,0.4051404529147678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,3072,0.25219999419318306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,2560,0.2218115594651964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,65536,11.268036736382378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,1536,0.1366728941599528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,2560,0.2210079961352878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,2048,0.07634488741556804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,2048,0.18270311090681288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,1024,0.09596977631251018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,1536,0.06198933389451769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,1536,0.1435297727584839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,2048,0.1793431176079644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,768,0.07118044296900432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,1024,0.05369066529803806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,1536,0.1419644488228692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,1024,0.10891555415259467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,512,0.05208177698983086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,1024,0.11638755268520778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,256,0.037498666180504694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,512,0.043735110097461276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,768,0.09031822284062703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,512,0.07451911105049981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,128,0.03015822172164917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,256,0.06467466884189181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,256,0.03913066784540812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,256,0.05794577466117012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,64,0.02534399926662445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,128,0.03914755582809448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,8192,32,0.025386666258176167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,128,0.06007111072540283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,128,0.04353600078158908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,64,0.061848892105950244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,32,0.06431377596325345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,768,0.04692711101637947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,768,0.08617777956856622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,512,0.07216088639365302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,65536,2.129812452528212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,51200,1.6456658045450847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,51200,3.7862951490614147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,16384,1.3878755569458008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,16384,0.44611109627617734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,65536,5.259751213921441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,16384,1.9117919074164496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,12288,1.0506329006618924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,16384,1.4322018093532984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,51200,4.993507385253906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,12288,0.35969866646660703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,12288,1.0694231457180448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,10240,0.9203955332438151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,10240,0.2502133316463894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,12288,1.0294355816311307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,51200,6.622912089029948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,8192,0.20588444338904488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,10240,0.9605787065294055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,65536,6.9414621988932295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,10240,0.8465733528137207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,8192,0.619384871588813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,7168,0.2088222238752577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,7168,0.5586337513393825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,8192,0.6517030927870009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,6144,0.3975493378109402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,7168,0.5348986519707574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,6144,0.16181333859761557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,7168,0.5115911165873209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,5120,0.13629244433508977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,5120,0.4046186606089274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,6144,0.4842791027492947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,5120,0.3795528941684299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,8192,0.6630986531575521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,4096,0.3083831204308404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,5120,0.3664524555206299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,4096,0.11799911657969157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,3584,0.2795555591583252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,4096,0.3557075659434001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,4096,0.2970844374762641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,3584,0.10672266615761651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,3072,0.22798310385810006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,3584,0.2863813241322835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,3072,0.09429333607355754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,2560,0.19219999843173555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,65536,9.63851843939887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,3072,0.23917865753173828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,3072,0.2277733219994439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,2048,0.14917066362169054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,2560,0.08835555447472467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,2560,0.20122755898369682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,1536,0.12168799506293403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,2048,0.07209689087337919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,2560,0.1952142185635037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,1024,0.08357244729995728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,2048,0.16214844915601942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,1536,0.05762310822804769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,1536,0.12866222858428955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,768,0.07081511285569933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,1536,0.1346400048997667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,1024,0.0476817786693573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,1024,0.09606311056349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,1024,0.09862311018837823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,512,0.05663288964165581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,768,0.07755111323462592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,768,0.08368977573182847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,256,0.03915199968549941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,512,0.03870222303602431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,512,0.062462223900689014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,512,0.06664178106519911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,128,0.026591110560629103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,256,0.03601866629388597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,2048,0.16412888632880315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,256,0.05191377798716227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,128,0.051368001434538096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,128,0.036115556955337524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,128,0.03937066594759623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,32,0.023448000351587932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,768,0.043004443248113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,32,0.0522400008307563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,3584,0.2705920007493761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,256,0.054642667373021446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,7168,64,0.021872000561820135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,64,0.05141866538259718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,6144,0.5752693282233344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,65536,1.7589235305786133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,51200,1.3862364027235243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,51200,3.1500852372911243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,16384,1.2820089128282335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,65536,4.1598125033908415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,16384,0.36015110545688206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,12288,0.8261120054456922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,16384,1.461873796251085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,16384,1.2456347147623699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,12288,0.2669484350416395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,51200,4.109252505832248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,65536,5.471570756700303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,12288,0.8588275379604764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,10240,0.734623114267985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,10240,0.2832346757253011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,51200,5.482169681125217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,10240,0.6951084666781955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,8192,0.5835999912685818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,10240,0.6926630867852105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,8192,0.1896053287718031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,12288,0.8604817920260959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,7168,0.3874826696183946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,7168,0.1577191087934706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,8192,0.8160106870863172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,6144,0.36099200778537327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,8192,0.5562773280673557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,7168,0.5846195750766331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,6144,0.18659555912017822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,5120,0.29197510083516437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,65536,7.965801662868923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,6144,0.4155235555436876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,6144,0.3860319985283746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,5120,0.12339199913872613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,4096,0.25389332241482204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,5120,0.3636835681067572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,3584,0.22586133744981554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,4096,0.3066186639997694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,5120,0.32509689860873753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,4096,0.2619413269890679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,3072,0.1776968903011746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,3584,0.09269777933756511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,3584,0.29419467184278697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,3584,0.23522755834791395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,2560,0.1543360021379259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,3072,0.08089422517352633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,3072,0.20957867304484049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,3072,0.20059733920627168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,2048,0.13722488615247938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,2560,0.07442844576305814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,2560,0.2200782166586982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,7168,0.4623342090182834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,1536,0.10137689113616943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,2560,0.17151555750105116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,2048,0.14740088250901964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,1024,0.07159733110004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,1536,0.051227556334601514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,2048,0.14356356196933323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,1536,0.11293689409891765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,768,0.05702310800552368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,1536,0.11651821931203206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,1024,0.04303644431961907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,4096,0.10023822387059529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,512,0.043306665288077466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,768,0.03903733359442817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,768,0.06813066535525851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,256,0.028593778610229492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,768,0.07277511225806342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,512,0.03566133313708835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,512,0.05920000208748711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,128,0.020909332566791113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,256,0.03319733341534933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,256,0.04782844583193461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,256,0.046033776468700834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,64,0.018421333697107103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,128,0.03278311093648275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,2048,0.0611048870616489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,6144,32,0.01885688967174954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,128,0.04478311207559374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,128,0.03558933403756883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,64,0.04468533396720886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,32,0.04530310961935255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,1024,0.08425333102544148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,1024,0.08697688579559326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,512,0.054495112763510815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,65536,1.3908560011121962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,51200,1.0779600143432617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,16384,1.025144894917806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,51200,2.937887191772461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,16384,1.2489555146959093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,65536,3.7587021721733938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,16384,0.2985173331366645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,16384,1.0242533153957791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,12288,0.712066650390625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,51200,3.5959705776638455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,12288,0.27189866701761883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,51200,4.488087972005208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,12288,0.6998986668056912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,10240,0.5704319741990831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,10240,0.1940257814195421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,12288,0.7517111036512586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,8192,0.48656267589992946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,10240,0.5839075512356228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,65536,5.904655032687717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,10240,0.6416968769497341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,8192,0.4495529068840875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,8192,0.19691199726528594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,7168,0.3980906539493137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,7168,0.13947289519839817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,8192,0.47803465525309247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,6144,0.3203279972076416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,65536,4.550369686550565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,7168,0.4177573257022434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,6144,0.12239733007219102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,7168,0.42670488357543945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,5120,0.10782133208380805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,5120,0.27532977528042263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,4096,0.20642844835917154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,6144,0.33356711599561906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,5120,0.2819306585523817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,3584,0.1790684461593628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,5120,0.28058576583862305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,4096,0.2696097691853841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,3072,0.16280355718400744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,4096,0.2311555544535319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,3584,0.20222132735782197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,3584,0.20251644982231987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,2560,0.11301510863833958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,3072,0.07255999909506904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,3072,0.1805217795901828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,2560,0.06333955791261461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,3072,0.181987550523546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,2560,0.14877155092027453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,2560,0.15292621983422175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,6144,0.3788168960147434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,1536,0.08006844255659315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,2048,0.05388799972004361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,2048,0.12294667296939427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,2048,0.12836266888512507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,1024,0.0596382220586141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,4096,0.10298666689130996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,1536,0.04553066690762838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,1536,0.09707555505964492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,768,0.04511911008093092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,1024,0.037979556454552546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,3584,0.09076799949010213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,1024,0.072744886080424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,512,0.034252444903055825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,1024,0.07744977871576945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,768,0.03490044342146979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,768,0.059077335728539355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,256,0.022496889034907024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,768,0.0651457773314582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,512,0.032131556007597185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,512,0.04696444339222378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,128,0.01795111099878947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,512,0.05275555451711019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,2048,0.11539733409881592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,256,0.02954044606950548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,256,0.0405982236067454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,64,0.01713511182202233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,5120,32,0.01846222248342302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,256,0.04077422287729051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,128,0.03844177722930908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,128,0.031077331966824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,32,0.03827910953097873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,1536,0.10222755538092719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,128,0.029161777761247423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,64,0.038013332419925265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,65536,1.0891884697808158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,51200,0.8745031356811523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,16384,0.9457048839992948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,51200,2.433526145087348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,65536,3.1409778594970703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,16384,0.7891813384162055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,16384,0.9385075039333768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,65536,3.3143662346733938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,51200,3.4337821536593967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,65536,4.221062978108724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,51200,2.607020484076606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,12288,0.1978071133295695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,16384,0.2711057662963867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,10240,0.4093040095435248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,12288,0.5869110955132378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,10240,0.16084710756937662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,12288,0.5428933567470974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,8192,0.3309662342071533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,8192,0.12891733646392822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,10240,0.48239999347262913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,7168,0.28942754533555776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,12288,0.6047049098544651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,10240,0.4676559766133626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,8192,0.44822486241658527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,8192,0.3697013325161404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,7168,0.13678399721781412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,7168,0.3366559876335992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,6144,0.26849155955844456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,7168,0.3235742251078288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,5120,0.22336355845133463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,6144,0.2850995593600803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,4096,0.1591271162033081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,5120,0.08865600162082249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,6144,0.2823191218905979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,5120,0.23725422223409018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,5120,0.25933067003885907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,3584,0.15169510576460096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,4096,0.07446666558583577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,4096,0.18965778085920545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,3072,0.1205937729941474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,3584,0.06775822242101033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,3584,0.16642044650183782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,4096,0.2375262313418918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,2560,0.11039821969138251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,3584,0.17538132932451037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,3072,0.062089780966440834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,3072,0.14556177457173666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,2048,0.08283911148707072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,3072,0.1528257793850369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,2560,0.1263902187347412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,2560,0.054604444238874644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,6144,0.10382311211691962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,1536,0.06232355700598823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,2048,0.04671288861168755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,2048,0.10285422537061904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,2560,0.13002667162153456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,1024,0.044779555665122144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,1536,0.03996444410747952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,2048,0.11131821738349067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,768,0.03513155712021722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,1024,0.033237334754731916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,1536,0.0868871079550849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,512,0.024262222978803847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,1024,0.06093155675464206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,1024,0.06688977612389459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,768,0.04961244596375359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,512,0.026529777381155226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,768,0.05650044149822659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,512,0.0389066669676039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,512,0.04569866591029697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,256,0.02455377744303809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,256,0.033590220742755465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,256,0.03509777784347534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,64,0.013033777475357056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,128,0.023751111494170293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,128,0.03156177865134345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,32,0.013410666750537025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,128,0.0269795556863149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,64,0.03102311160829332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,32,0.031067556805080835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,1536,0.08152355750401814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,768,0.029866665601730347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,256,0.017687999539905123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,4096,128,0.014135999812020196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,65536,0.9937697516547309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,51200,0.7783644464280871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,16384,0.7982106738620334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,65536,2.663498772515191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,51200,2.366699642605252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,16384,0.7995857662624783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,16384,0.22164800431993273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,12288,0.5460391044616699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,51200,2.4052675035264754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,16384,0.7298844655354818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,51200,2.956312815348307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,12288,0.17236622174580893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,65536,3.069122738308377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,65536,3.685966067843967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,10240,0.42130133840772843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,12288,0.5245511266920302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,10240,0.1499626636505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,8192,0.3017813364664714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,8192,0.11785599920484756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,10240,0.42199288474188906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,7168,0.2612711058722602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,8192,0.3410009013281928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,10240,0.474256886376275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,7168,0.10705244541168213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,6144,0.2210622231165568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,8192,0.4360204537709554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,7168,0.3711564540863037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,5120,0.1896524429321289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,7168,0.3062195512983534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,6144,0.2521582179599338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,12288,0.49049690034654403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,4096,0.14696621894836426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,5120,0.22210400634341768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,5120,0.22552445199754503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,3584,0.1442168951034546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,4096,0.18201155132717559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,3584,0.15196888976626927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,3584,0.16471021705203587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,3584,0.06313421991136339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,3072,0.11085244682100083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,6144,0.09398577610651652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,6144,0.27525777286953396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,3072,0.0554986662334866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,5120,0.09380444553163315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,3072,0.13135111331939697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,3072,0.14362043804592556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,2560,0.10677599906921387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,2048,0.0830942193667094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,2560,0.05005422234535217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,4096,0.16961067252688936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,2560,0.11196445094214545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,1536,0.061816891034444175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,2048,0.04258133305443657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,2048,0.09278044435713027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,1024,0.04548266530036926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,1536,0.0359226663907369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,2048,0.10244533088472153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,1536,0.0731840001212226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,1536,0.08221155405044556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,768,0.03433333171738519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,1024,0.029370665550231934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,1024,0.05491377909978231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,512,0.025891555680169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,1024,0.06240266561508179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,768,0.026724444495307073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,768,0.04458222124311659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,256,0.017670222454600863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,512,0.02383822202682495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,512,0.03497599893146091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,512,0.0429111123085022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,4096,0.06788177622689141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,128,0.014166222678290473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,256,0.02156177825397915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,256,0.029820445511076186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,256,0.03182044294145372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,64,0.01606399979856279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,2560,0.12235732873280843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,128,0.02104088829623328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,128,0.028247998820410833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,128,0.024367999699380662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,64,0.027666666441493567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,32,0.027439999911520217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,768,0.05197688937187195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3584,32,0.017135999268955655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,65536,0.8827937973870171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,65536,1.9326648712158203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,16384,0.6685377756754557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,51200,0.6560071309407552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,51200,1.8947395748562281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,16384,0.2158355580435859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,12288,0.45744890636867946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,16384,0.6051848729451498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,51200,2.3874551984998917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,16384,0.6311270925733778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,65536,2.732415941026476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,51200,2.106485366821289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,10240,0.3539244333902995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,12288,0.5212248696221246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,65536,3.3917901780870228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,12288,0.47637510299682617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,10240,0.14361600081125894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,8192,0.28771999147203237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,10240,0.38628445731268984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,8192,0.29536088307698566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,7168,0.22687111960517037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,10240,0.40761423110961914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,7168,0.09428177939520942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,6144,0.1954195499420166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,8192,0.3171146710713704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,7168,0.2952231036292182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,7168,0.27455467647976345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,5120,0.16898577743106416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,6144,0.08581333027945624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,12288,0.15504444970024958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,6144,0.22859644889831543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,4096,0.13166577286190456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,5120,0.07260977559619479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,5120,0.19184445010291207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,6144,0.23552799224853516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,3584,0.11650399367014568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,4096,0.06268711222542657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,4096,0.1504488918516371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,5120,0.19951466719309488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,3072,0.10598044262992011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,4096,0.16562310854593912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,3584,0.05733333031336466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,3584,0.13431910673777261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,2560,0.07974933253394233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,8192,0.10685866408877903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,3584,0.14690044191148546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,3072,0.05081511206097073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,3072,0.1162373357348972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,2560,0.04553955462243822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,3072,0.12798133161332872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,2560,0.10166844394471909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,2048,0.038345777326160006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,2560,0.11031999852922227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,2048,0.08241955439249675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,1024,0.034703999757766724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,2048,0.09280533260769314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,1536,0.06485955582724677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,1536,0.03159644537501865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,768,0.026949332820044622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,1536,0.07430133554670545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,1024,0.02605244517326355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,1024,0.04861333303981357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,512,0.01906488835811615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,1024,0.05614844295713636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,768,0.02315555512905121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,768,0.03922577699025472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,256,0.01407911048995124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,768,0.0471955570909712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,512,0.021078222327762183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,512,0.030973331795798406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,128,0.011723555624485016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,512,0.03900977638032701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,256,0.019128888845443726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,2048,0.06464355521731906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,256,0.02650044361750285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,64,0.010584000084135266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,256,0.0295688874191708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,32,0.011220444407727985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,3072,1536,0.05223377876811557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,128,0.01899377836121453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,128,0.025044444534513686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,64,0.024383111132515803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,32,0.02457422183619605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,65536,0.7429457770453559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,51200,1.2543101840549045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,128,0.021959111094474792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,51200,0.5904613071017796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,65536,1.9941902160644531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,16384,0.44925509558783633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,16384,0.5023502243889703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,16384,0.2191386620203654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,12288,0.3673857847849528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,51200,2.0068783230251737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,65536,2.3679021199544272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,10240,0.27126577165391713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,12288,0.17589955859714082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,16384,0.5864791340298122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,12288,0.37699198722839355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,65536,3.0453156365288625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,10240,0.12268355157640244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,8192,0.22852532068888345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,12288,0.47570843166775173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,10240,0.33325688044230145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,51200,1.7754400041368272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,10240,0.3500320116678874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,8192,0.25484977828131783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,7168,0.09102133247587417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,8192,0.2900079886118571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,7168,0.2268142170376248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,5120,0.1456773281097412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,6144,0.1944728824827406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,6144,0.0757840010854933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,5120,0.06700799862543742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,5120,0.16372444894578722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,5120,0.17715822325812447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,7168,0.18900533517201742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,4096,0.13180443975660536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,4096,0.05578311284383138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,3584,0.10505421956380208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,4096,0.1460568904876709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,6144,0.1639768944846259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,3072,0.08283378018273248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,3584,0.050273779365751475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,7168,0.24232355753580728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,3584,0.11679200331370036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,3584,0.13128444883558485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,2560,0.06691644589106242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,3072,0.04507377743721008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,3072,0.10209066338009304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,6144,0.21687289079030356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,2048,0.058223108450571694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,2560,0.04011111127005683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,2560,0.08722755644056533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,2560,0.09848444329367743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,1536,0.0439991123146481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,2048,0.03330755564901564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,2048,0.07269866598976983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,8192,0.09642489088906182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,2048,0.0830142233106825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,1536,0.026879999372694228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,1536,0.057245333989461265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,768,0.022808889547983806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,1536,0.06658399767345853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,1024,0.04240533378389147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,1024,0.05077422327465481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,512,0.016639111770523917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,768,0.020283554991086323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,768,0.03476444549030728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,768,0.042753779225879245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,256,0.012289777398109436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,512,0.027280888623661462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,512,0.018911111685964797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,3072,0.11361332734425862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,256,0.023197333017985027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,128,0.009983110758993361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,256,0.025059555967648823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,1024,0.02980977627966139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,128,0.01753866672515869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,128,0.0217848883734809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,64,0.008924444516499838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,1024,0.022249778111775715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,32,0.008993777963850234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,64,0.021083555287784998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,32,0.020950223008791607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2560,4096,0.10760977533128525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,256,0.017507554756270517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,128,0.01997599999109904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,65536,0.6425395541720921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,65536,1.5744132995605469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,51200,0.9017368952433268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,512,0.0336426662074195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,16384,0.3928933408525255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,51200,0.5597644382052952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,16384,0.17375644048055014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,16384,0.43990667661031085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,65536,2.2962000105116105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,51200,1.5690497292412653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,65536,2.009311040242513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,12288,0.2899528874291314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,16384,0.46052180396185977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,12288,0.13299466503991023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,10240,0.21628888448079428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,12288,0.3400675455729167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,8192,0.1737217770682441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,10240,0.1055377788013882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,12288,0.34937156571282285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,10240,0.2927030987209744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,10240,0.3194062179989285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,7168,0.15720178021325007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,8192,0.08847200208240086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,8192,0.23116887940300834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,6144,0.133952882554796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,8192,0.24059645334879556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,7168,0.07460177607006498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,7168,0.19295822249518502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,5120,0.10801955726411606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,7168,0.21255911721123588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,6144,0.06568355692757501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,6144,0.1794817712571886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,4096,0.08645777569876777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,5120,0.05746044715245565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,5120,0.1420231130388048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,3584,0.07371199793285794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,5120,0.15536888440450033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,4096,0.11259110768636067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,4096,0.13042755921681723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,3072,0.0722000002861023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,3584,0.04365066687266032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,3584,0.0997919970088535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,51200,1.5283768971761067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,2560,0.0560471084382799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,3584,0.11496178309122722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,3072,0.039134220944510564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,2048,0.04812888966666328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,3072,0.08694666624069214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,2560,0.03364177876048618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,2560,0.07476177480485705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,1536,0.0351182222366333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,2560,0.08590488963656956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,2048,0.028175112273957994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,2048,0.061724444230397545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,1024,0.02466933263672723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,2048,0.07240355677074857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,1536,0.023909333679411147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,6144,0.18513066238827178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,1536,0.049422221051322095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,768,0.018975999620225694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,1536,0.05748799774381849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,1024,0.02000977761215634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,4096,0.049492445256974965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,1024,0.036229332288106285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,512,0.013772444592581855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,1024,0.04436444573932224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,768,0.0185244447655148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,256,0.010399111443095738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,768,0.029848890172110662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,512,0.016859556237856548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,128,0.008650666309727563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,768,0.036714666419559054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,512,0.02342044479317135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,256,0.01626755628320906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,512,0.02948088778389825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,64,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,256,0.019985778464211356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,2048,32,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,128,0.01587377819750044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,256,0.02308711078431871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,128,0.018564444449212816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,3072,0.09928266869650947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,128,0.018592889110247295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,64,0.01793244481086731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,32,0.017887999614079792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,65536,0.5262577798631456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,51200,0.7261368963453504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,65536,1.1853119532267253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,16384,0.28623467021518284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,51200,0.4333484437730577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,51200,1.1819707022772896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,16384,0.1412008868323432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,65536,1.6426684061686199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,51200,1.304310268825955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,12288,0.19212978416019014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,16384,0.45628801981608075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,65536,1.721111085679796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,16384,0.47046486536661786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,10240,0.16376354959275988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,12288,0.10590133402082656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,8192,0.14066489537556967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,10240,0.09194933043585883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,10240,0.23184445169236925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,12288,0.3240453402201335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,7168,0.12433066633012559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,8192,0.1869279940923055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,10240,0.2700631088680691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,6144,0.10419288608762954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,7168,0.06403199831644694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,8192,0.23411467340257433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,7168,0.16221600108676487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,5120,0.08451555834876166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,6144,0.057504885726504855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,7168,0.19293155935075548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,6144,0.13808533880445692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,4096,0.05961866511238945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,5120,0.050118221176995166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,6144,0.16711110538906523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,5120,0.116230222913954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,3584,0.058633777830335826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,12288,0.276094224717882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,4096,0.043079998758104115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,5120,0.14327288998497859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,4096,0.09327466620339288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,3072,0.04972355564435323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,3584,0.038352890147103205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,3584,0.08226133055157132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,3072,0.03289422061708238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,8192,0.07125955820083618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,3584,0.10352710882822673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,3072,0.07187555233637492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,2048,0.036361777120166354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,2560,0.027687999937269423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,3072,0.09071377913157146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,2560,0.06156444549560547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,1536,0.02590933276547326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,2048,0.02423200011253357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,2560,0.07749155494901869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,2048,0.051113777690463595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,1024,0.018738667170206707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,2048,0.0648533304532369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,1536,0.04057688845528497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,1536,0.021428444319301184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,768,0.01516533394654592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,1536,0.05225333240297106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,1024,0.01832088828086853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,512,0.011400000088744693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,1024,0.030048890246285334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,768,0.01695999999841054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,768,0.02496266696188185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,256,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,512,0.015629332926538255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,512,0.01979111135005951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,4096,0.1162266731262207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,128,0.008567999634477828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,2560,0.043100443151262074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,256,0.0148044443792767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,256,0.01683999929163191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,64,0.008076444268226624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,256,0.020745777421527438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,128,0.01480977733929952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1536,32,0.007499555746714274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,128,0.015501333607567681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,128,0.016357334123717416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,64,0.0147706667582194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,32,0.014503111441930136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,1024,0.0392595562669966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,65536,0.689085324605306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,51200,0.4954542054070367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,768,0.032877332634396024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,65536,0.42542844348483616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,512,0.02661777701642778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,16384,0.18271377351548937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,16384,0.108916441599528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,65536,1.1820266511705186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,16384,0.2855822245279948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,12288,0.13589510652754042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,16384,0.34720622168646914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,51200,0.8856924374898275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,65536,1.3561111026340062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,10240,0.11491289403703477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,12288,0.22028888596428764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,12288,0.267081790500217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,10240,0.0769066678153144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,51200,1.088732401529948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,10240,0.18504977226257324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,10240,0.22568978203667533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,8192,0.1420960028966268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,7168,0.07958488994174533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,8192,0.18056978119744194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,51200,0.3323688771989611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,6144,0.07242933246824476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,7168,0.05272622240914238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,7168,0.12548800309499106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,7168,0.16015288564893934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,6144,0.04672266708479988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,5120,0.057187557220458984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,12288,0.08473155233595107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,6144,0.10906933413611518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,4096,0.04344800114631653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,8192,0.0817048880789015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,5120,0.04114044374889798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,5120,0.09171199798583984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,5120,0.11816622151268853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,4096,0.03468177715937296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,4096,0.0745119982295566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,8192,0.05880622069040934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,4096,0.09785155455271403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,3072,0.0328382220533159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,3584,0.029370665550231934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,3584,0.06546133094363742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,2560,0.027492443720499676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,3072,0.02643022272321913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,3072,0.057736886872185596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,3072,0.07491111093097262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,2560,0.022434666752815247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,2560,0.06430044439103869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,6144,0.1396239995956421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,2048,0.04123377799987793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,1536,0.018636445204416912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,2048,0.020074667202101815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,3584,0.04223111271858215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,2048,0.05446133348676893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,1024,0.013576000101036496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,1536,0.01796711153454251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,1536,0.032631112469567194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,768,0.011348444554540845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,1536,0.04327911138534546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,1024,0.015807110402319167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,1024,0.02422399984465705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,3584,0.08604622549480861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,512,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,768,0.014782221780882942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,768,0.02030044462945726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,256,0.006500444478458828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,768,0.027666666441493567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,512,0.013759999639458127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,512,0.016188444362746347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,2560,0.048947556151284106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,128,0.005969777703285217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,512,0.022503111097547743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,256,0.013406222065289816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,64,0.005882666756709416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,128,0.01309866706530253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,32,0.0058782220714622075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,128,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,128,0.014129777749379476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,64,0.011703111231327057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,32,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,65536,0.5333671040005153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,1024,0.032285332679748535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,65536,0.4097839991251628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,1024,2048,0.02353688908947839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,256,0.013071111506885953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,51200,0.44696002536349827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,256,0.01719199948840671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,65536,0.9892515606350369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,51200,0.31970667839050293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,16384,0.10772089163462321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,51200,0.7517786555820041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,12288,0.10698933071560329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,16384,0.3126720057593452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,65536,1.210185792711046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,10240,0.10127288765377468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,12288,0.08630488978491889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,12288,0.18949422571394178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,10240,0.15831644005245632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,10240,0.07671466800901625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,8192,0.07103555732303195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,12288,0.24514577123853895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,51200,0.9597413804796008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,7168,0.06639377938376533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,10240,0.2015697823630439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,8192,0.057208889060550265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,6144,0.05553866757286919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,8192,0.12566844622294107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,16384,0.14083288775549993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,7168,0.11111644903818767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,5120,0.04445688923199972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,6144,0.045891556474897593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,6144,0.0960657795270284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,6144,0.12559466891818577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,5120,0.0403511126836141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,16384,0.310444434483846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,5120,0.10646666420830621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,4096,0.06547555658552381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,4096,0.08760799964269002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,4096,0.033365332418017916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,3584,0.030396444929970637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,7168,0.05189511179924011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,8192,0.16430755456288657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,3584,0.057760000228881836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,3072,0.02684266699684991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,3584,0.028914666838116113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,7168,0.1440826654434204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,2560,0.02294844388961792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,3584,0.07758044534259372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,3072,0.024914667010307312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,3072,0.05070844292640686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,4096,0.034845332304636635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,3072,0.06752000252405803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,2560,0.022171555293930903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,2560,0.043601777818467885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,1536,0.014893333117167154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,2560,0.05822755230797661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,2048,0.019658666517999437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,2048,0.03615022367901272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,1024,0.011700444751315646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,2048,0.048840890328089394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,1536,0.017606221967273288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,1536,0.028716444969177246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,768,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,1536,0.038880000511805214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,1024,0.015822221835454304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,1024,0.021595555875036452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,512,0.0069982219073507525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,1024,0.029191109869215224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,768,0.014782221780882942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,768,0.017958222164048087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,256,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,768,0.02493600050608317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,512,0.013738666971524557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,512,0.014449778530332776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,128,0.004962666581074397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,512,0.021698666943444148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,256,0.013103110922707452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,256,0.011798222031858234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,64,0.006529777828190062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,32,0.004556444370084339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,256,0.016230222251680162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,5120,0.08106399907006158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,128,0.010339555641015371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,768,2048,0.019631110959582858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,128,0.013334222137928009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,64,0.01000533335738712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,768,32,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,65536,0.39885955386691624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,51200,0.3122311168246799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,65536,0.31853066550360787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,16384,0.101928002304501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,51200,0.2627120018005371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,65536,0.8064666854010688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,51200,0.6287671195136176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,12288,0.07421066363652547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,16384,0.08808177709579468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,16384,0.2945342328813341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,128,0.013077333569526672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,10240,0.05905955367618137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,65536,1.0589947170681424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,12288,0.06908710797627766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,12288,0.15634132756127253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,10240,0.060032890902625195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,12288,0.20858932865990532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,10240,0.13036799430847168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,7168,0.04124622212515937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,10240,0.17947822146945527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,8192,0.10554399755265977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,8192,0.04561333192719353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,6144,0.03676711188422309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,7168,0.04138755467202928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,7168,0.09316355652279323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,7168,0.12776354948679605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,16384,0.20470221837361655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,6144,0.08072000079684787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,5120,0.030633777379989624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,6144,0.03692622317208184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,4096,0.023472888602150813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,6144,0.11182222101423477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,5120,0.032179554303487144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,5120,0.06801066795984904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,51200,0.8583724233839246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,3584,0.020818667279349435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,4096,0.024677332904603746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,5120,0.0946719977590773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,8192,0.046804444657431714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,4096,0.055440889464484326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,3072,0.018592000007629395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,3584,0.021661332911915247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,4096,0.07842666572994657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,2560,0.016872889465755887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,3072,0.019479110836982727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,3584,0.04916888806555006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,2048,0.013799111048380533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,3072,0.0428711109691196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,3584,0.06883022520277235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,3072,0.06013155645794339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,2560,0.03661244445376926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,1536,0.011085333095656501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,2048,0.016032889485359192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,2048,0.030557334423065186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,2048,0.04286666711171468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,1024,0.008273777862389883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,1536,0.014500444134076437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,1536,0.024401777320437964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,1536,0.03409333361519708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,768,0.007058666812049017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,1024,0.013111111190583972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,1024,0.018351111147138808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,1024,0.025916443930731878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,512,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,768,0.012728888955381183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,768,0.01551822159025404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,768,0.02235466738541921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,256,0.004589333302444882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,512,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,512,0.012551111479600271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,2560,0.017479111750920612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,512,0.018742221924993727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,128,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,256,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,8192,0.14387288358476427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,256,0.010060444474220276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,2560,0.0517759985393948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,64,0.003519111209445529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,512,32,0.0037297776175869834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,256,0.014688889185587565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,128,0.01146311147345437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,128,0.008978666530715095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,128,0.011738667057620155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,64,0.008466666771305932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,65536,0.24448977576361763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,512,32,0.008191111187140146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,51200,0.1850524478488498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,16384,0.061025778452555336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,65536,0.3130871189965142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,51200,0.24855465359157988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,65536,0.6646409034729004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,16384,0.1756053368250529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,51200,0.5184399816724989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,16384,0.2357546753353543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,65536,0.9023155636257596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,12288,0.06632533338334826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,10240,0.037393778562545776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,51200,0.7322213384840223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,8192,0.030633777379989624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,12288,0.18392711215549043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,10240,0.059232890605926514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,10240,0.11072088612450494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,7168,0.026446221603287592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,8192,0.04460888769891527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,8192,0.12576177385118273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,7168,0.0787351131439209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,7168,0.04021599888801575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,6144,0.022535110513369244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,16384,0.08297777838177152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,6144,0.03569777806599935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,6144,0.0679804417822096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,5120,0.016912889149453905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,6144,0.09730488724178737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,12288,0.13093777497609457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,4096,0.013783999615245394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,5120,0.03093866507212321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,5120,0.05719111363093058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,5120,0.08268533150355022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,10240,0.15301422278086343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,3584,0.012763555679056378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,4096,0.02398577829202016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,8192,0.0903973314497206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,3072,0.01108888867828581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,4096,0.067358222272661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,3584,0.021254221598307293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,3584,0.04111822115050422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,3072,0.0196079992585712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,3072,0.0362097786532508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,12288,0.0467564430501726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,3072,0.05138577686415779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,2048,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,2560,0.01757866640885671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,2560,0.031247112486097548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,2560,0.04408355553944906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,1536,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,2048,0.015944000747468736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,2048,0.02566755480236477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,2048,0.03609955641958449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,1024,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,1536,0.020639111598332722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,1536,0.014695111248228284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,4096,0.046572443511750966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,1536,0.029719998439153034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,768,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,1024,0.013441777891582914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,1024,0.015630222029156156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,512,0.004214222232500712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,1024,0.0225600004196167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,768,0.01274400038851632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,768,0.013185777597957186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,2560,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,768,0.019832889238993328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,256,0.003444444388151169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,7168,0.11144088374243842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,512,0.011744889120260874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,512,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,128,0.003142222141226133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,512,0.016544888416926067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,256,0.011352889239788055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,256,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,256,0.013073777986897362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,128,0.011356444822417365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,32,0.0030782222747802734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,128,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,64,0.006677333265542984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,32,0.006679111056857639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,65536,0.18114044931199816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,51200,0.14003111256493464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,65536,0.32102754380967885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,65536,0.5938595665825738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,256,64,0.0028044444819291434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,16384,0.04439022143681844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,256,128,0.0070462218589252895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,51200,0.4930586814880371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,51200,0.25217066870795357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,65536,0.8859573470221626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,12288,0.035960889524883695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,16384,0.08464266856511433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,16384,0.15435289012061224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,10240,0.02933599882655674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,12288,0.11720889144473606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,51200,0.6926097869873047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,12288,0.06758222315046522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,16384,0.23112800386216906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,8192,0.02479377720091078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,12288,0.17699021763271758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,10240,0.05920800235536364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,3584,0.05899466408623589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,7168,0.021372444099850122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,10240,0.09879910945892334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,8192,0.07977333333757189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,6144,0.016978666186332703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,7168,0.041323555840386286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,8192,0.12326221995883518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,5120,0.011389333340856763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,7168,0.07010400295257568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,7168,0.1090808908144633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,6144,0.03683377636803521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,6144,0.060552000999450684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,4096,0.009299555586444007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,6144,0.09556266996595596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,5120,0.030947556098302204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,5120,0.05067288875579834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,3584,0.008623999853928884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,5120,0.08101689153247409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,4096,0.041433778074052595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,4096,0.024283554818895128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,3072,0.007582222421964009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,4096,0.06530577606625028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,3584,0.021264889174037512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,3584,0.036610666248533465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,2560,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,3584,0.05748977926042345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,3072,0.01984444426165687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,3072,0.03197866678237915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,2048,0.0058373332851462895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,2560,0.018153778380817838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,3072,0.05086311035686069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,2560,0.027456000447273254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,1536,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,10240,0.15059110853407118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,2048,0.016420443852742512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,2048,0.02310133311483595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,1024,0.004190222256713443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,1536,0.014941333068741692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,2048,0.035240888595581055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,768,0.003543111185232798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,1536,0.028972443607118394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,1024,0.013603555659453074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,1024,0.014131555954615274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,512,0.0032195556494924757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,1024,0.022126222650210064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,768,0.012950222525331708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,768,0.01203733351495531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,256,0.0026542221506436667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,768,0.01923022170861562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,512,0.012063110868136087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,512,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,128,0.002537777854336633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,512,0.016154666741689045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,256,0.011406222151385413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,256,0.007558222446176741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,64,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,128,0.011386666860845355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,128,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,128,32,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,64,0.00598933340774642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,2560,0.04318222072389391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,65536,0.17104000515407988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,32,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,128,1536,0.018570666511853535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,51200,0.13062577777438694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,16384,0.04830044507980347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,12288,0.03404888841840956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,16384,0.14705067210727268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,10240,0.02829866607983907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,65536,0.561356438530816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,51200,0.4407004515329997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,12288,0.11028089125951131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,7168,0.020990222692489624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,256,0.012773333324326409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,10240,0.09185066488054063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,6144,0.016199999385409884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,8192,0.07438755697674222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,5120,0.010022222167915767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,4096,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,7168,0.06539911031723022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,3584,0.007342221836249034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,6144,0.05669599771499634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,3072,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,4096,0.039108445247014366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,2560,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,3584,0.0347013341055976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,2048,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,3072,0.030256887276967365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,1536,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,2560,0.025905778010686237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,1024,0.003927111211750242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,2048,0.021996445126003687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,768,0.0034844444857703317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,8192,0.023984889189402264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,1536,0.017699556218253244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,512,0.0031519999934567344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,256,0.0025315555847353404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,1024,0.013722666435771518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,8192,0.0455991095966763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,128,0.002234666711754269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,768,0.011628444823953839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,32,0.002159111201763153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,512,0.009334222310119206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,256,0.007104000283612146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,128,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,64,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,5120,0.04763466782040066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,64,32,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,16384,0.04493511054250929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,51200,0.12568267186482748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,12288,0.034648001194000244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,16384,0.14777866999308267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,10240,0.02700177828470866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,8192,0.023786667320463393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,51200,0.4396328926086426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,12288,0.10991022321912979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,65536,0.5609973271687826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,7168,0.01963733302222358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,6144,0.014731556177139282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,10240,0.09190222289827134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,5120,0.009759999811649323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,8192,0.07460711399714152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,4096,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,64,64,0.0021137777302000257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,3584,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,3072,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,7168,0.06541155444251166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,6144,0.05685510900285509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,2560,0.0053093334039052325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,2048,0.004599110947714912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,5120,0.04771022333039177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,4096,0.03918755716747708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,1536,0.004851555658711327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,3584,0.03657866517702738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,1024,0.0038373333712418876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,768,0.0034764446318149567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,2560,0.02681955529583825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,512,0.002871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,2048,0.022674666510687932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,256,0.00249422217408816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,1024,0.013944000005722046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,768,0.011765333513418833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,128,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,64,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,512,0.009681777821646797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,32,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,256,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,128,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,64,0.00565244464410676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,32,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8192,32,65536,0.17367288801405165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,3072,0.03223022156291538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8192,32,1536,0.018588445252842374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,16384,1.9400950537787542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,12288,4.771656036376953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,16384,5.84148449367947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,16384,6.758196512858073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,12288,1.3587289386325414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,12288,4.524311913384332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,51200,7.429303487141927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,10240,1.2236710654364693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,10240,3.777545081244575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,8192,0.9024995168050131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,8192,3.010159174601237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,10240,3.5590621100531687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,8192,2.8937867482503257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,7168,0.7947644657558866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,6144,0.8311137623257108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,7168,3.0252034929063587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,6144,2.023876402113173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,7168,2.570311016506619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,6144,2.12229241265191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,5120,0.5830373234219021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,4096,0.4887404441833496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,4096,1.5579955842759874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,5120,1.8627342647976344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,5120,1.7306381861368816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,4096,1.4572426478068035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,3584,0.428272008895874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,3584,1.297441800435384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,3072,1.1251298056708443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,3072,0.3758542272779677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,3584,1.2070506413777669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,2560,0.3790915542178684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,2560,0.9736702177259656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,2048,0.6907546785142687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,2560,0.8998764885796441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,128,0.009668444593747457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,2048,0.3014435503217909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,2048,0.7409164640638563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,1536,0.612502204047309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,1536,0.22439821561177573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,3072,1.0923990673489041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,1024,0.3583564493391249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,1536,0.5612906879848903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,768,0.3015742301940918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,1024,0.20423289140065512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,512,0.2319262292650011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,1024,0.4348311159345839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,768,0.1782239940431383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,51200,18.878280639648438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,512,0.15018488301171198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,256,0.1647146675321791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,768,0.3290710979037815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,128,0.13811111450195312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,256,0.14818844530317518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,64,0.10191822052001953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,65536,128,0.14145155747731528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,65536,32,0.11370044284396702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,256,0.25518666373358834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,128,0.243248012330797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,32,0.2515840000576443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,64,0.25375111897786456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,512,0.2979377905527751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,16384,4.283853318956163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,51200,5.846294403076172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,65536,51200,27.626661512586807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,65536,7.352345784505208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,16384,1.4278639687432184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,16384,5.135070376925998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,12288,1.0366897583007812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,12288,3.3948461744520397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,12288,3.624298519558377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,10240,0.9166533152262369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,10240,2.9365244971381297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,51200,13.763778686523438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,10240,2.9782674577501087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,8192,0.6816568904452853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,8192,2.7059830559624563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,8192,2.2884445190429688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,7168,1.980655034383138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,7168,1.9924888610839844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,7168,0.6294319894578722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,6144,1.7737680011325414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,6144,0.563185797797309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,65536,17.333548651801216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,6144,1.6853502061631944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,5120,1.4693181779649522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,5120,0.47120708889431423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,4096,0.38979021708170575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,4096,1.2374977535671658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,5120,1.3655679490831163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,3584,0.9523235956827799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,3584,0.3484124342600505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,3584,0.9651181962754992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,4096,1.0610284805297852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,3072,0.29979554812113446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,3072,0.901225725809733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,2560,0.259390221701728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,3072,0.8079599804348416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,2560,0.7276310920715332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,2048,0.2167982260386149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,2560,0.6994009017944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,2048,0.5776462025112575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,1536,0.4041617711385091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,1536,0.1747431092792087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,1024,0.28026845720079213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,2048,0.6347297562493218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,1024,0.16075021690792507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,768,0.21513333585527208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,1536,0.6370622316996256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,512,0.17484266228146025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,768,0.12327378325992161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,1024,0.44259021017286515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,256,0.11873155170016819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,768,0.3230355580647786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,512,0.1217155588997735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,128,0.11122043927510579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,512,0.22643200556437174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,256,0.09559822082519531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,64,0.08233155806859334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,256,0.2170177830590142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,51200,128,0.09786222378412883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,51200,32,0.08414222134484185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,128,0.19059822294447157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,64,0.19477600521511504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,32,0.20379911528693306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,51200,22.714925130208332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,65536,2.350320816040039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,16384,1.3742169274224176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,51200,1.9698827531602647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,16384,0.44512801700168186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,51200,4.390185038248698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,16384,1.7418764962090387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,12288,0.34690133730570477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,12288,1.3544532987806532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,65536,5.8964123196072045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,12288,1.1602107153998482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,10240,0.31522488594055176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,10240,1.1090293460422092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,8192,0.23689066039191353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,10240,0.9417573081122504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,8192,0.7860097885131836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,8192,0.729298644595676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,7168,0.21295289198557535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,7168,0.8249182171291776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,6144,0.1866675615310669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,7168,0.6435742378234863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,6144,0.5953582127888998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,5120,0.4404275682237413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,51200,7.411895751953125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,6144,0.5327377849155003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,4096,0.13046933545006648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,4096,0.3359368907080756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,3584,0.27349334292941624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,5120,0.45888355043199325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,4096,0.3802328904469808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,3584,0.1157288948694865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,3072,0.23635021845499674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,3072,0.10601066880755955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,3584,0.31389954355027944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,2560,0.23035200436909994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,3072,0.2733866638607449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,2048,0.17456177870432535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,2560,0.24281067318386504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,2048,0.07759377691480848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,1536,0.1337546639972263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,5120,0.15987200207180446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,2048,0.19281510512034097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,1024,0.09661244683795506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,1536,0.06305955515967475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,768,0.07214222351710002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,1024,0.0517119997077518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,1536,0.1498826609717475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,512,0.054859555429882474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,768,0.04824355575773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,256,0.03828088773621453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,768,0.0894506639904446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,512,0.04283377859327528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,128,0.02808800008561876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,512,0.07710578044255574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,256,0.03829866647720337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,64,0.02494222256872389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,128,0.0381004446082645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,128,0.06525688701205783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,16384,32,0.02607822252644433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,16384,2560,0.09308889177110459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,51200,65536,31.069875081380207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,32,0.06794399685329862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,1024,0.11235555013020833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,65536,9.787809583875868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,256,0.06984711355633207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,16384,64,0.06643911202748616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,16384,1.0378026962280273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,65536,1.7814292907714844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,51200,1.3263253106011283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,16384,0.37873421774970156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,16384,1.2379653718736436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,12288,0.2655404408772786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,12288,1.1793271170722115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,10240,0.7222302224900988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,51200,3.3462062411838107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,12288,0.8826515409681531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,10240,0.26101599799262154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,65536,4.122873094346788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,8192,0.17851200368669298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,10240,0.7214426464504666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,8192,0.6367280218336318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,7168,0.15965688228607178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,8192,0.5702426698472765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,7168,0.4761813481648763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,6144,0.40288710594177246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,6144,0.13831822077433267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,5120,0.29634666442871094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,7168,0.491846243540446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,5120,0.11859022246466742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,6144,0.487335999806722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,4096,0.23655554983350965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,5120,0.3947048981984456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,4096,0.09879200326071845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,3584,0.2251324388715956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,3584,0.08918578094906277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,3072,0.20074578126271567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,3584,0.26122045516967773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,3072,0.07944622304704455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,2560,0.16430221663581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,3072,0.2223306761847602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,2048,0.14087733957502577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,2560,0.07002577516767713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,2560,0.18029422230190703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,1536,0.10862221982744004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,2048,0.06860444280836317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,2048,0.1463413370980157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,1024,0.07255999909506904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,51200,5.837988535563151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,768,0.059142225318484835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,1536,0.11510222487979466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,1536,0.049499554766549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,512,0.04428889022933113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,1024,0.08743733167648315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,768,0.038233776887257896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,4096,0.27076178126864964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,256,0.028767112228605483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,768,0.07637155718273587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,512,0.034908443689346313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,128,0.02343822187847561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,256,0.03154755632082621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,512,0.05963289075427585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,64,0.01984355515903897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,12288,32,0.020475554797384474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,128,0.03130666746033563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,128,0.05048000150256687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,64,0.05141244331995646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,32,0.05336977707015144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,12288,1024,0.04298044575585259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,65536,7.274553934733073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,12288,256,0.053344001372655235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,65536,1.455842653910319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,51200,1.1055848863389757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,16384,1.0120230780707467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,16384,0.30609599749247235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,16384,0.9996684392293295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,12288,0.22632622718811035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,12288,0.7789831161499023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,51200,2.8419022030300565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,10240,0.5065733591715494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,10240,0.23959289656745064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,65536,3.405255211724175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,12288,0.9444559945000542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,8192,0.46057600445217556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,10240,0.605679988861084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,8192,0.16608799828423396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,8192,0.5026053322686089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,7168,0.13816000355614558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,7168,0.38861955536736387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,6144,0.3510302172766791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,6144,0.11966310607062446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,7168,0.39934754371643066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,5120,0.27549423111809623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,6144,0.3570426570044623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,5120,0.10540977451536392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,4096,0.2150275574790107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,4096,0.08873955408732097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,5120,0.35923110114203555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,3584,0.1973724497689141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,4096,0.22669598791334364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,3584,0.07926311095555623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,3072,0.16453866163889566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,2560,0.12775999969906277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,3072,0.07093155384063721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,3584,0.20793422063191733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,2560,0.060544888178507485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,3072,0.18198577562967935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,2048,0.11224088403913711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,1536,0.09202400181028579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,51200,4.871529897054036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,2560,0.15359289116329616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,1024,0.059213333659701876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,1536,0.04288711150487264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,2048,0.13083022170596653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,1536,0.09961155388090347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,1024,0.03679822219742669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,768,0.03407999873161316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,1024,0.07415378093719482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,256,0.024692444337738886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,512,0.031184001101387873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,512,0.05124622252252367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,256,0.028221332364612158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,128,0.02160355614291297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,256,0.04633422361479866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,65536,6.061478932698567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,64,0.01791466606987847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,128,0.027451554934183758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,32,0.0181697772608863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,128,0.04314755400021871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,64,0.046273777882258095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,10240,2048,0.05177866750293308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,768,0.04547022117508782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,32,0.044849776559405856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,10240,512,0.033801777495278254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,10240,768,0.05958844555748833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,16384,0.6316133075290257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,65536,1.3589039908515081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,51200,0.9840382470024956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,16384,0.2776515483856201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,12288,0.19935822486877441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,16384,0.8209048906962076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,51200,2.2794905768500433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,12288,0.6774177551269531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,12288,0.6658826404147679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,10240,0.4927075703938802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,10240,0.1752177741792467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,8192,0.40655822224087185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,65536,3.0651893615722656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,8192,0.17676356103685167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,10240,0.47238757875230575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,7168,0.3271244366963704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,8192,0.3857937653859456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,7168,0.12920267052120632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,6144,0.2894195450676812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,6144,0.12018222279018825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,5120,0.23450401094224718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,6144,0.27816979090372723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,7168,0.438460456000434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,51200,3.4244374169243703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,5120,0.09787644280327691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,4096,0.18091821670532227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,5120,0.2439671092563205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,3584,0.16124178303612605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,4096,0.07476088735792372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,3072,0.131750226020813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,3584,0.07708089219199286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,2560,0.12053510877821182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,4096,0.1996471087137858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,3072,0.06282933553059895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,3584,0.16884533564249674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,2560,0.06172800064086914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,3072,0.14389422204759386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,1536,0.07027200195524427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,2560,0.14290222856733534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,2048,0.04536622100406223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,1024,0.04820266697141859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,768,0.03763022356563144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,2048,0.10815111133787368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,1536,0.03692266676161025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,1024,0.032356444332334734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,512,0.0277173337009218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,1536,0.08224711153242323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,256,0.020085333122147452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,1024,0.06311200062433879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,768,0.02823466724819607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,128,0.015477332803938123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,512,0.025552888711293537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,768,0.04991022083494398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,65536,4.693495008680555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,64,0.013407111167907715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,256,0.02314666741424137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,512,0.041723555988735624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,32,0.013735111388895245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,8192,128,0.023209777143266466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,256,0.036701334847344294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,128,0.034565332863065935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,32,0.03430577781465318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,8192,64,0.03637333379851447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,8192,2048,0.0977697769800822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,16384,0.6368080245123969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,65536,1.1246550877888997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,51200,0.878786669837104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,16384,0.6999200185139974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,16384,0.2594960000779894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,12288,0.17843466334872773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,10240,0.4279759989844428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,12288,0.5106648868984646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,51200,2.094448937310113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,10240,0.17613689104715982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,12288,0.755791982014974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,8192,0.3599066734313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,10240,0.41632623142666286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,65536,2.530178705851237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,8192,0.1276879972881741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,7168,0.300439993540446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,7168,0.10712889167997573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,6144,0.26368710729810924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,6144,0.10423111253314549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,5120,0.1870746612548828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,7168,0.3971502251095242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,6144,0.2630551126268175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,4096,0.14255289236704508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,5120,0.08436799711651272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,4096,0.06726221905814277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,3584,0.13520356019337973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,5120,0.2972968949211968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,3072,0.11956800354851617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,3584,0.06031022469202677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,3584,0.14892978138393825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,3072,0.05310488740603129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,3072,0.14300444391038683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,65536,4.061444600423177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,8192,0.4514337645636664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,2048,0.0794008904033237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,2560,0.10973599884245132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,2560,0.04723377691374885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,1536,0.06865511337916057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,2048,0.04363022247950236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,1024,0.04296088880962796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,2048,0.09057155582639907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,1536,0.07356622483995226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,768,0.035262223747041486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,1024,0.027549333042568628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,4096,0.22238400247361925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,512,0.023059555225902136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,768,0.024910221497217815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,768,0.04431110951635572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,2560,0.10094666481018066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,512,0.022435555855433147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,512,0.03604177633921305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,128,0.014687110980351767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,256,0.021232000655598108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,256,0.03122577733463711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,64,0.012383999923865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,128,0.020592888196309406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,32,0.012401777837011548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,128,0.03011288907792833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,51200,2.917909410264757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,7168,1536,0.03291822142071194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,64,0.0299635562631819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,32,0.03020711077584161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,7168,1024,0.05592444207933214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,7168,256,0.01771022213829888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,16384,0.45620356665717227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,51200,0.728723578982883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,65536,0.9428986443413628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,16384,0.19928710990481904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,16384,0.6090168952941895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,12288,0.15010844336615667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,12288,0.538548469543457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,10240,0.3150871064927843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,51200,1.7672480477227106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,12288,0.5035804642571343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,10240,0.1700542238023546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,65536,2.2525929345024958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,10240,0.3759511046939426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,8192,0.10314666562610203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,7168,0.30775732464260525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,7168,0.09116533067491318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,6144,0.21153510941399467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,7168,0.3206293318006727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,5120,0.16845777299669054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,6144,0.08146578073501587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,51200,2.569443596733941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,4096,0.1405537790722317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,5120,0.06986755794949003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,6144,0.3017253345913357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,3584,0.11919377909766303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,4096,0.061568889353010386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,4096,0.14459289444817439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,65536,3.3645981682671438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,3584,0.05741155809826321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,3584,0.1476773288514879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,8192,0.2666169007619222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,2560,0.08954044183095296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,3072,0.05120177732573616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,2048,0.06659466690487333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,3072,0.1118311087290446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,2560,0.04082489013671875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,1536,0.056919111145867236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,2560,0.09620710876252915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,2048,0.03625955515437656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,1024,0.034753776258892484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,2048,0.08931289116541545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,1536,0.029614223374260798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,768,0.02906222144762675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,1536,0.06281866629918416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,1024,0.02538577715555827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,512,0.02123022245036231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,1024,0.04842488964398702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,768,0.022295999858114455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,256,0.015507555670208402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,512,0.020077332854270935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,768,0.03970844546953837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,128,0.014308444327778287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,512,0.03168977631462945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,256,0.018238221605618794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,5120,0.18552888764275444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,64,0.014743111199802823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,32,0.015290665957662793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,6144,128,0.01789688898457421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,6144,3072,0.09949688778983222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,128,0.026167111264334783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,64,0.02607822252644433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,32,0.025851555996470984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,8192,0.3225146664513482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,65536,0.7486604584587945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,51200,0.59970490137736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,6144,256,0.02731288969516754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,51200,1.2860071394178603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,16384,0.16958488358391655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,16384,0.4818728764851888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,12288,0.33722400665283203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,12288,0.15885421964857313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,65536,2.012607150607639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,16384,0.6584755579630533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,10240,0.2663946681552463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,12288,0.3758213255140517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,10240,0.13307021723853216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,8192,0.08912355369991726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,51200,2.2784303029378257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,7168,0.20825333065456816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,10240,0.37067288822597927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,8192,0.2762604554494222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,6144,0.18013332949744332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,65536,2.8493885464138455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,7168,0.21806844075520834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,6144,0.1898204485575358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,4096,0.1114764478471544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,5120,0.06410577562120226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,3584,0.10389510790506999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,4096,0.051164444949891835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,5120,0.15559289190504286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,4096,0.12627289030287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,3072,0.08134933312733968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,3584,0.044827553961012095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,8192,0.2437208758460151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,2560,0.07267644670274523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,3072,0.04026755690574646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,3584,0.11015111207962036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,2048,0.05754577451282077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,3072,0.09706666734483506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,2560,0.03537688983811273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,2560,0.0830506682395935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,2048,0.030788444810443457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,7168,0.08496089114083184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,1024,0.03147555390993754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,2048,0.06975555419921875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,1536,0.02555911077393426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,6144,0.07026311424043444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,1536,0.054783112472958036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,512,0.016952888833151925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,1024,0.041441778341929116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,768,0.019321777754359774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,768,0.033963554435306124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,512,0.017567111386193168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,512,0.027614222632514104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,128,0.011056000159846412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,256,0.016117332710160148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,256,0.023352889551056757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,64,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,128,0.01648355523745219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,128,0.022655111220147874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,1536,0.042915556165907115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,64,0.022687110635969374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,5120,32,0.022408000297016565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,768,0.0247653325398763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,5120,0.1451128853691949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,5120,1024,0.02180355621708764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,256,0.01292888902955585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,16384,0.3421440124511719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,51200,0.47814400990804035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,65536,0.6092711024814182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,5120,32,0.009752888646390703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,51200,1.0494462119208443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,16384,0.14332711696624756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,65536,1.3638266457451715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,12288,0.3164435492621528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,10240,0.24481688605414498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,16384,0.42798932393391925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,10240,0.11009599765141805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,8192,0.2069093386332194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,10240,0.27959465980529785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,8192,0.07965155442555745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,7168,0.1818702220916748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,51200,1.6526702245076497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,8192,0.2052053345574273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,7168,0.07185777690675524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,6144,0.15704354974958631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,65536,2.2422684563530817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,6144,0.05854577488369412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,5120,0.11517155170440674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,7168,0.18463377157847086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,4096,0.0921928882598877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,6144,0.1615333292219374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,5120,0.057393776045905225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,4096,0.042797333664364286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,5120,0.16802755991617838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,3072,0.07028088967005412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,4096,0.10926133394241333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,3584,0.0387733346886105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,2560,0.05485777722464668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,3584,0.09397688839170669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,3072,0.033442666133244835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,2048,0.04861688945028517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,12288,0.30229422781202525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,2560,0.029998222986857098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,3072,0.08996444278293186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,1536,0.036306666003333196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,2048,0.02595022155178918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,1024,0.023610666394233704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,2048,0.05765333440568712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,1536,0.02236533330546485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,768,0.019528888993793063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,1024,0.018615111708641052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,1024,0.03468622101677789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,512,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,768,0.01687466601530711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,768,0.028985778490702312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,256,0.010696888797812991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,512,0.01550044450494978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,3584,0.07744266589482625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,512,0.02272266646226247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,128,0.010025777750545079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,256,0.014495111174053617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,256,0.019707555572191875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,64,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,12288,0.11982044908735488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,4096,32,0.009647111097971598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,128,0.019254222512245178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,2560,0.07115555471844144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,64,0.01885422236389584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,32,0.01886577738655938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,4096,1536,0.0466284453868866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,65536,0.5686337682935927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,51200,0.4417573346032037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,65536,1.125308460659451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,51200,0.846009784274631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,4096,128,0.01479200025399526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,16384,0.317483557595147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,12288,0.24069865544637045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,16384,0.13155910703870985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,12288,0.10102222363154094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,10240,0.23267467816670737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,16384,0.37560354338751895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,10240,0.09023733271492852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,8192,0.16172088517083064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,12288,0.2704382207658556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,51200,1.4374755223592122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,7168,0.13023200300004747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,8192,0.06880444288253784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,10240,0.29020799530877006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,65536,1.802795622083876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,7168,0.06105155415005154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,6144,0.12756978140936956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,5120,0.10376177893744574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,6144,0.05293777916166517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,7168,0.16303200191921657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,4096,0.08429600132836236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,5120,0.04564533299869961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,3584,0.06617599725723267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,6144,0.14204266336229113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,4096,0.038763556215498186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,5120,0.12249422073364258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,3584,0.03381777803103129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,2560,0.046451555358039014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,4096,0.09669599930445354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,3584,0.09654577573140462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,3072,0.030332446098327637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,2560,0.027480888697836135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,1536,0.03091199861632453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,2560,0.06263822317123413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,2048,0.023530667026837666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,1024,0.02330222229162852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,2048,0.053800000084771045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,1536,0.019726221760114033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,768,0.019138667318556044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,1536,0.042246222496032715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,1024,0.016872889465755887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,1024,0.03149866726663377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,768,0.015839111473825242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,768,0.0259226659933726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,8192,0.2631395657857259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,256,0.011063111325105032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,512,0.014698666003015308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,128,0.00830488900343577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,512,0.020648888415760465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,256,0.013744889034165276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,3072,0.05561866362889608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,64,0.007248000138335758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,256,0.0181351105372111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,32,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3584,128,0.013754666679435305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,128,0.01753422286775377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,3072,0.07445777787102593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,64,0.01690933273898231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3584,32,0.01698933376206292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,65536,0.4573209020826552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,51200,0.37424800131056046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,51200,0.771358225080702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,65536,1.035254266526964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,16384,0.29562399122450084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,12288,0.21175289154052734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,16384,0.1132391161388821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,2048,0.04051822092798021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,16384,0.4016879929436578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,10240,0.1888915565278795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,12288,0.09019377496507432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,12288,0.29099376996358234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,65536,1.630416022406684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,8192,0.15250933170318604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,10240,0.07505777809354994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,10240,0.20219377676645914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,8192,0.06261599726147123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3584,512,0.01517777807182736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,7168,0.05353600117895338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,8192,0.1640391084882948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,5120,0.08802844418419732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,7168,0.14562133948008218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,6144,0.04820444517665439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,4096,0.07130933470196195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,6144,0.12247555785708958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,5120,0.04058666692839728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,3584,0.05916000074810452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,5120,0.10359022352430557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,4096,0.032871110571755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,3072,0.05175644490453932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,4096,0.08505866924921672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,3584,0.029993779129452173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,51200,1.1623173819647896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,2560,0.042839109897613525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,3584,0.07512266768349542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,3072,0.026963555150561865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,2048,0.03489511211713155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,3072,0.06504978073967828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,2560,0.024540444215138752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,1536,0.027453333139419556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,2048,0.021427555216683283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,7168,0.1339226696226332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,2560,0.05576622486114502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,1024,0.019076445036464267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,768,0.015092444088723926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,2048,0.04730755421850416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,1536,0.018761777215533786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,512,0.011249778171380361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,1024,0.01586933268441094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,1536,0.03780533207787408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,256,0.008436444732877944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,768,0.014929778046078153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,1024,0.027842667367723253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,512,0.013787555197874704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,128,0.007608888877762689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,512,0.0188737776544359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,64,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,256,0.016315556234783597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,128,0.012798222402731577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,32,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,128,0.015562666787041558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,64,0.015211555692884656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,32,0.015198222464985318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,65536,0.4308426645067003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,3072,6144,0.11246666643354629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,51200,0.6394320064120823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,3072,768,0.023523555861579046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,16384,0.25280711385938853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,65536,1.064992904663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,3072,256,0.01297511077589459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,51200,0.34536533885531956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,12288,0.19696533679962158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,16384,0.10308621989356147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,12288,0.08067822456359863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,65536,1.239040904574924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,51200,0.9691564771864148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,10240,0.16780355241563585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,16384,0.29308001200358075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,10240,0.06942311260435316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,12288,0.2544248898824056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,7168,0.11894044611189102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,10240,0.18971733252207437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,6144,0.10112266408072577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,7168,0.04817600051561991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,8192,0.14970221784379747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,5120,0.07939822143978544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,7168,0.12571467293633357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,6144,0.04399999976158142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,4096,0.061837335427602134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,5120,0.03583644496070014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,3584,0.0566177765528361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,6144,0.10992177989747788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,5120,0.0916373332341512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,4096,0.02979466650221083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,3072,0.051390220721562706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,3584,0.027202667461501226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,2560,0.040476444694730974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,4096,0.07362844546635945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,3072,0.024517332514127094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,2048,0.03385688861211141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,3584,0.0650728874736362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,2560,0.0218542218208313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,3072,0.05680711401833428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,1536,0.027253333065244887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,2048,0.019278221660190158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,2560,0.048884444766574435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,1536,0.016938666502634685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,2048,0.041420443190468684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,768,0.01315822203954061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,1024,0.014839111102951897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,1536,0.03398044572936164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,8192,0.1341057750913832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,512,0.01073333372672399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,1024,0.02495288848876953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,8192,0.05670666694641113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,256,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,768,0.014085332552591959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,768,0.02088800072669983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,512,0.012829333543777466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,128,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,64,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,256,0.01238666640387641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,512,0.017318222257826064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,32,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,256,0.014431110686726041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,128,0.014105778601434497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,64,0.01346577786737018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2560,32,0.013763555222087435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2560,1024,0.019641778535313077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,65536,0.3381768862406413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,65536,0.6830471356709799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,51200,0.5825004577636719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,51200,0.26416089799669057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,16384,0.10417511065800984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,65536,1.010764439900716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,51200,0.7862515449523926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,16384,0.2997359964582655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2560,128,0.012094222009181976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,10240,0.12049333254496257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,12288,0.06715377834108141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,10240,0.05776266919242012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,8192,0.08865511417388916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,12288,0.20399910873836943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,7168,0.07264977693557739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,10240,0.15291644467247859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,8192,0.04504444532924228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,6144,0.06297155883577135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,7168,0.04082311193148295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,5120,0.05449777841567993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,7168,0.109497778945499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,4096,0.04718133476045397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,6144,0.09269689189063178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,5120,0.031416889694001936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,16384,0.1806124448776245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,5120,0.07841866546207003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,4096,0.02557955516709222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,3072,0.03455555438995361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,3584,0.02348711093266805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,4096,0.06279289060168795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,3584,0.05539466606246101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,12288,0.13763288656870523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,3072,0.021493333909246657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,2048,0.024640000528759424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,2560,0.019521777828534443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,3072,0.04867555697758993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,1536,0.019301333361201815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,2560,0.04197600152757433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,2048,0.017263111140992906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,1024,0.014119999276267158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,2048,0.036093334356943764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,1536,0.015475556254386902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,768,0.011366222467687396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,1536,0.028945777151319716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,8192,0.12191200256347656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,1024,0.01347733371787601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,512,0.00853777759604984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,6144,0.034952001439200506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,1024,0.021879111727078755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,256,0.0063582224150498705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,768,0.018273777431911893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,512,0.011758222348160215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,3584,0.038916445440716214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,128,0.006181333214044571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,512,0.0148053334818946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,64,0.005691555639108022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,256,0.012772444221708508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,32,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,128,0.011390222443474663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,128,0.01238044434123569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,2048,2560,0.029682666063308716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,64,0.011767110890812345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,2048,32,0.011899555722872416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,65536,0.30769422319200307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,51200,0.4550906817118327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,65536,0.5984515613979763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,51200,0.21880443890889487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,16384,0.14328444004058838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,12288,0.10435466633902656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,16384,0.08740711212158203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,768,0.012454222473833295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,65536,0.8046275774637858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,51200,0.6287804179721409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,12288,0.06376800272199842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,16384,0.2410213417477078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,2048,256,0.011071110765139261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,8192,0.07317777474721272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,12288,0.16134756141238743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,10240,0.0574222207069397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,8192,0.04446844591034783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,6144,0.05664800273047554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,7168,0.03856177793608771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,8192,0.10110311375723945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,7168,0.08923911386066014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,6144,0.03433333171738519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,6144,0.07751911216311984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,5120,0.028766221470303003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,4096,0.03516533308558994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,5120,0.06503466765085857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,3584,0.030696001317765977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,4096,0.024255999260478552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,4096,0.05379466546906365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,3072,0.026935999592145283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,3584,0.021910222040282354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,3584,0.04775733417934842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,2560,0.022610666023360357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,3072,0.019167999426523846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,7168,0.07139377461539374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,3072,0.042100446091757886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,2048,0.01872711049185859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,10240,0.12500177489386663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,1536,0.015398222539159985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,2560,0.036222222778532244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,2048,0.016380444169044495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,5120,0.04193688763512505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,1024,0.010788444843557147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,1536,0.013780444032616086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,2048,0.030320889419979517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,768,0.00867377801073922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,512,0.006917333437336816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,1536,0.02479644450876448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,1024,0.012134222520722283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,256,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,1024,0.018784888916545443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,512,0.010763555765151978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,768,0.01573244399494595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,128,0.005200888961553574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,512,0.013073777986897362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,64,0.004625777817434735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,256,0.011128000087208219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,128,0.010231999887360467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,32,0.005205333232879639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1536,10240,0.09471733040279812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,64,0.01033688916100396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,32,0.010055999788973067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,2560,0.018410666121376883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,65536,0.3708213435278998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,65536,0.22194222609202066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,51200,0.28826843367682564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,16384,0.09608533647325303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,51200,0.18921421633826363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,65536,0.6083848741319444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,768,0.011414222419261932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,12288,0.07347822189331055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,16384,0.0600151088502672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,10240,0.06425511174731784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,16384,0.15495110882653132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,51200,0.48204533259073895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1536,256,0.010148444109492833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,12288,0.04744355546103584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,8192,0.04882311158710056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,7168,0.04422400063938565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,10240,0.0443173348903656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,12288,0.11940622329711914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,8192,0.030988444884618122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,6144,0.038329776790406965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,10240,0.10059911012649536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,7168,0.027095999982621934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,8192,0.08053600125842623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,6144,0.024247111545668706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,7168,0.07065600156784058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,4096,0.02347822156217363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,3584,0.02115999990039402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,5120,0.02065333392884996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,6144,0.06162399715847439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,3072,0.01886044442653656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,5120,0.0518142216735416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,3584,0.016581333345837064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,4096,0.042837331692377724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,2560,0.016359999775886536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,3584,0.038402666648228966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,3072,0.015485333071814643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,2048,0.013887999786270989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,3072,0.033835556772020124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,2560,0.014153778553009033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,1536,0.010896888871987661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,2560,0.029164443413416546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,2048,0.013008000122176277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,1024,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,2048,0.024570667081409033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,768,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,1536,0.02016799979739719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,1024,0.010833777487277985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,512,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,1024,0.015765332513385348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1536,128,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,768,0.010348444183667501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,768,0.01315199997689989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,5120,0.031641778018739485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,512,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,128,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,512,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,256,0.009343999955389235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,256,0.009339555270142024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,4096,0.01789955629242791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,32,0.0038968887594011095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,128,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,64,0.00833511104186376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,1024,32,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,65536,0.31051821178860134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,1536,0.011767110890812345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,51200,0.2518826590643989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,65536,0.21995555029975045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,16384,0.07883911000357734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,65536,0.526293330722385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,51200,0.18649510542551676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,51200,0.4113973246680365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,64,0.003839111162556542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,16384,0.05873599979612562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,1024,128,0.009334222310119206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,10240,0.05125244458516439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,12288,0.04682666725582547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,8192,0.03875111209021674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,10240,0.041401777002546526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,12288,0.10258044136895074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,10240,0.08653333452012803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,7168,0.035409778356552124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,8192,0.030202666918436687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,6144,0.02760533326201969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,7168,0.027162666122118633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,8192,0.07003821929295857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,5120,0.024597333537207708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,7168,0.061349332332611084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,6144,0.023676445086797077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,4096,0.019912888606389362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,5120,0.02053777707947625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,6144,0.05344444513320923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,1024,256,0.004837333328194088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,3584,0.01684177749686771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,3072,0.014530667000346713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,5120,0.04530222217241923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,4096,0.03765155540572272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,3584,0.016571554872724745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,2560,0.012433778080675336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,12288,0.059346669250064425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,3584,0.033407999409569636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,3072,0.015414221419228448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,2048,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,3072,0.029618667231665716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,16384,0.13494043880038792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,2560,0.01427288850148519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,1024,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,2048,0.013109332985348172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,2560,0.025719111164410908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,768,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,1536,0.01184800018866857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,2048,0.021726222501860723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,512,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,1536,0.017976888351970248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,768,0.01036622209681405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,1024,0.013797333670987023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,256,0.0038444445365005066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,768,0.011700444751315646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,128,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,512,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,256,0.009369778136412302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,64,0.003170666595300039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,256,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,32,0.003169777699642711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,128,0.00775911079512702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,4096,0.017768000562985737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,64,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,768,32,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,65536,0.2210702233844333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,768,1536,0.008912888665994009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,51200,0.1764533387290107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,16384,0.05571733580695259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,51200,0.13627644379933676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,1024,0.010882666541470421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,51200,0.3560791015625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,512,0.009872000250551436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,16384,0.11433955033620198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,16384,0.04680622286266751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,10240,0.0330062210559845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,12288,0.0377475553088718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,768,128,0.009341333475377824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,10240,0.033611555894215904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,10240,0.07326400279998779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,7168,0.022602667411168415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,8192,0.02387111054526435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,8192,0.05886489152908325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,7168,0.020624889267815482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,6144,0.019070222973823547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,65536,0.17187910609775117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,7168,0.052321778403388135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,5120,0.016192888220151264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,6144,0.018447111050287884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,4096,0.01348533315791024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,6144,0.045569777488708496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,5120,0.01629866659641266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,65536,0.4372417661878798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,3584,0.012246222131782107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,5120,0.03854399919509888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,12288,0.04312800036536323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,4096,0.014610666367742749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,4096,0.03195555673705207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,3584,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,2560,0.009190222455395592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,3584,0.02882400155067444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,3072,0.01276000009642707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,8192,0.026713778575261433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,3072,0.02527911133236355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,2560,0.012067555553383298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,1536,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,2560,0.022073777185546026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,2048,0.011362666885058085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,1024,0.0052826665341854095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,2048,0.018886221779717337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,1536,0.010670222342014313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,768,0.004588444613748127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,1536,0.01572444372706943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,1024,0.009934222532643212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,1024,0.012110222544935016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,768,0.009367110828558603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,768,0.010390222072601318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,256,0.0032231110251612137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,512,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,3072,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,128,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,256,0.008651555412345463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,256,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,64,0.00282222218811512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,512,128,0.00869599978129069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,32,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,128,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,64,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,65536,0.17216266526116264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,32,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,51200,0.13419466548495823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,512,0.0038942222793896994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,65536,0.16864000426398384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,16384,0.043972445858849414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,65536,0.35657689306471085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,512,0.008984888593355814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,51200,0.13622932963901097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,12288,0.03468177715937296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,16384,0.04668622215588888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,51200,0.278677331076728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,16384,0.09427111016379462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,8192,0.020976889464590285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,12288,0.07122400071885851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,512,2048,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,7168,0.013778666655222574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,10240,0.059693336486816406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,8192,0.023766222927305434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,6144,0.012034667034943899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,7168,0.020337777005301583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,7168,0.0425946679380205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,5120,0.01036533299419615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,6144,0.018407111366589863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,6144,0.03765599926312765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,4096,0.0084906667470932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,5120,0.0166275550921758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,5120,0.032058666149775185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,3584,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,4096,0.014539556370841132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,4096,0.02645422187116411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,3072,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,10240,0.028969777954949275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,3584,0.013798221945762634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,12288,0.03770222266515096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,2560,0.006445333361625671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,3584,0.023702222439977858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,3072,0.012793777717484368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,512,12288,0.08733689122729832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,2048,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,1536,0.0048062221871482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,3072,0.02103999919361538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,2560,0.018550222118695576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,2048,0.011360000405046674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,1024,0.0039048890272776284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,8192,0.04838755395677355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,1536,0.010692444112565784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,2048,0.01593155496650272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,768,0.003560888684458203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,512,0.0031795555518733132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,1024,0.009769777456919352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,768,0.009326222042242685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,1024,0.010669333239396414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,256,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,768,0.00909511082702213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,512,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,128,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,512,0.008284444610277811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,256,0.008623110751310984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,128,0.008369777765538957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,32,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,128,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,64,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,65536,0.09772533178329468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,32,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,10240,0.03382488754060533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,256,2560,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,51200,0.07715021901660495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,65536,0.17247022522820366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,16384,0.030008888906902734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,1536,0.013211555778980255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,51200,0.13842044936286077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,12288,0.026480888326962788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,51200,0.2473226653205024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,16384,0.05076977941724989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,10240,0.016141333513789706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,16384,0.08289955721961127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,256,64,0.0024497778051429326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,12288,0.03938222262594435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,256,256,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,12288,0.06298577785491943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,10240,0.03517599900563558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,7168,0.011764444410800934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,8192,0.025290666355027094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,10240,0.052982221047083534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,6144,0.01036533299419615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,8192,0.043175998661253184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,7168,0.022441777918073866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,5120,0.00868711123863856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,7168,0.03812177644835578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,6144,0.019278221660190158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,4096,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,6144,0.03325244453218248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,5120,0.01750577820671929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,3584,0.00666222224632899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,4096,0.01556711064444648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,5120,0.028774221738179524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,3072,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,65536,0.3157706790500217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,4096,0.023738667368888855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,3584,0.014371555712487964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,2048,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,3072,0.013293333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,3584,0.021517333057191636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,1536,0.0038977778620190094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,2560,0.012596444123321109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,3072,0.019160888261265226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,1024,0.003522666792074839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,2048,0.011679111255539788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,1536,0.010724444356229572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,2048,0.01481599940194024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,768,0.0031582222630580268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,8192,0.013156444662147097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,1536,0.01237777786122428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,512,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,1024,0.009949333137936061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,768,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,256,0.0024666666156715816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,768,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,512,0.00904444439543618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,128,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,512,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,64,0.0021271111650599376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,256,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,128,0.008351999852392409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,32,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,128,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,64,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,32,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,65536,0.0901706682311164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,16384,0.029334220621320937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,51200,0.07271466652552287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,12288,0.02346666653951009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,128,2560,0.01684266659948561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,10240,0.014018666413095264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,16384,0.08278933498594496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,51200,0.2470853328704834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,1024,0.010011555420027839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,8192,0.012740444805887012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,7168,0.011101333631409539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,6144,0.009755555954244401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,12288,0.0630577802658081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,5120,0.008787555827034844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,10240,0.05305244525273641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4353,128,256,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,4096,0.007243555453088548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,7168,0.0376835564772288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,3584,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,3072,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,6144,0.03302222159173753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,2560,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,5120,0.028201778729756672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,2048,0.004592000196377437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,4096,0.02353777819209629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,3584,0.021294222937689886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,1536,0.004187555362780889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,3072,0.01891288823551602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,128,2560,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,768,0.003177777760558658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,2560,0.01682222220632765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,512,0.0028000000036425064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,2048,0.014671110444598727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,256,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,1536,0.01216533366176817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,1024,0.009870222045315636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,65536,0.3175404336717394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,128,0.0021173333128293357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,768,0.008642666869693333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,64,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,512,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,32,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,256,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,65536,0.08991911013921101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,128,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,64,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,51200,0.07066133287217882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,32,0.0053093334039052325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,16384,0.02667377723587884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,64,8192,0.04282133446799385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,12288,0.020206222931543987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,10240,0.01240355521440506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,8192,0.010301333334710863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,65536,0.31632622083028156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,16384,0.08323111136754353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,7168,0.009571555587980483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,51200,0.2473848925696479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,6144,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,10240,0.05307199888759189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,5120,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,64,1024,0.0035253332720862497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,8192,0.04276177618238661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,3584,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,7168,0.03792711098988851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,6144,0.033167110549079046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,5120,0.028175112273957994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,2560,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,4096,0.023601777023739282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,2048,0.00452888881166776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,3584,0.022112889422310725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,1536,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,3072,0.019498666127522785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,1024,0.0031848889258172777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,2560,0.01738577749994066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,2048,0.01478844384352366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,1536,0.012443555725945367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,512,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,1024,0.009962666365835402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,256,0.0021946666141351065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,768,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,128,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,12288,0.06275288926230536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,64,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,512,0.00756533361143536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,4096,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,256,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,128,0.005290666802061929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,64,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,3072,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4353,32,32,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4353,32,768,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,16384,1.875734117296007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,16384,5.244894239637587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,16384,6.369494967990452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,16384,5.710259755452474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,12288,4.4652383592393665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,51200,7.290859646267361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,12288,1.300829357571072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,12288,4.348140292697482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,10240,3.4594364166259766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,10240,1.1322506798638237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,12288,4.187121073404948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,10240,3.6274265713161893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,8192,0.8623218006557889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,8192,2.7516738043891062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,10240,3.2921085357666016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,8192,2.878923628065321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,8192,2.671303007337782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,7168,0.7763235304090711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,7168,2.809266620212131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,7168,2.33497068617079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,7168,2.416317409939236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,51200,17.313702053493923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,6144,0.6561813354492188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,6144,2.055532455444336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,6144,2.1343661414252386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,5120,1.6660302480061848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,5120,0.5804204410976833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,6144,2.0098408593071833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,5120,1.7257270812988281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,4096,1.3161564932929146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,4096,0.46084801355997723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,5120,1.6709519492255316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,4096,1.4060701794094508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,3584,0.40287288029988605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,3584,1.2953049341837566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,4096,1.397613313462999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,3584,1.1785137388441298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,3072,0.9710346857706705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,3072,0.36631377538045246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,3584,1.1529146830240886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,3072,1.0824818081325955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,2560,0.8179457982381185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,2560,0.3201848930782742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,3072,1.000388463338216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,2048,0.6621075736151801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,2560,0.8180249002244738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,2560,0.8666959868537055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,2048,0.3031217787000868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,2048,0.7323777940538195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,1536,0.5268088976542155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,1536,0.2176764408747355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,2048,0.7139164606730143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,1536,0.5484835306803385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,1024,0.4308115641276042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,1536,0.5202648904588487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,1024,0.17467378245459664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,51200,18.462844848632812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,1024,0.41562843322753906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,1024,0.40931023491753477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,768,0.3096506595611572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,512,0.22942667537265352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,768,0.3429253366258409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,768,0.18803021642896864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,768,0.3114071157243517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,256,0.1455679999457465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,512,0.1386444436179267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,512,0.2958257728152805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,128,0.11400532722473145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,512,0.28306934568617076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,256,0.11570844385359023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,256,0.32818489604526097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,65536,128,0.11549866199493408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,256,0.1966764397091336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,64,0.09937422143088447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,65536,128,0.15351200103759766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,128,0.24220177862379286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,65536,32,0.10739822520150079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,64,0.2436222235361735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,32,0.24841599994235566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,65536,51200,26.025853474934895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,51200,5.564467536078559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,65536,7.2746327718098955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,16384,4.9405670166015625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,51200,12.868553161621094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,16384,5.375816769070095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,16384,1.346554650200738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,65536,16.99523247612847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,12288,3.4445663028293185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,16384,4.484174092610677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,12288,0.9634595447116427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,12288,3.530112796359592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,51200,20.04963853624132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,12288,3.230178621080187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,10240,2.774874581231011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,10240,0.8205608791775174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,10240,2.9665234883626304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,10240,2.6470514933268228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,8192,0.6786950959099664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,8192,2.4082132975260415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,8192,2.139780468410916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,7168,1.8688187069363063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,8192,2.0845750172932944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,7168,0.6049128638373481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,7168,1.9456070793999567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,6144,1.588320838080512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,6144,0.5186399883694118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,6144,1.566292444864909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,7168,1.8176178402370875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,6144,1.5762702094184027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,65536,27.00504218207465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,5120,1.2545920477973092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,5120,1.3811857435438368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,4096,0.9935821957058377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,5120,1.3158293830023873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,5120,0.44460264841715497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,4096,0.36407023006015354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,4096,1.08805423312717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,4096,0.9868319829305013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,3584,0.3229120042588976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,3584,0.8990755081176758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,3584,0.934483528137207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,3072,0.8606062465243869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,3072,0.2960933314429389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,3584,0.8875946468777127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,2560,0.6619822184244791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,3072,0.8684222433302137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,2560,0.25655465655856663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,3072,0.7454711066351997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,2560,0.9892391628689237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,2048,0.21679111321767172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,2560,0.6283315552605523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,2048,0.6401217778523763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,2048,0.5797777705722386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,2048,0.5193973117404515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,1536,0.4083875550164117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,1536,0.215101334783766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,1024,0.26750487751430935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,1536,0.44327110714382595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,768,0.20741777949863008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,1024,0.1382417811287774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,1536,0.4087502161661784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,1024,0.35179999139573837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,512,0.16663111580742732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,768,0.12308711475796169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,1024,0.35338756773206925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,768,0.25949954986572266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,256,0.10114755233128865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,512,0.10806844631830852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,512,0.22473155127631295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,768,0.27039821942647296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,512,0.20276178254021537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,128,0.10027022494210137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,256,0.09245600303014119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,256,0.19834666781955293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,256,0.15723644362555608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,64,0.07524444659550984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,51200,128,0.09242577685250176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,51200,32,0.07835377587212457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,128,0.11935999658372666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,128,0.1875448889202542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,32,0.19602666960822213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,51200,64,0.19175733460320366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,65536,2.2356302473280167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,51200,4.303482479519315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,65536,5.524158053927952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,16384,1.3770987192789714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,51200,1.819226582845052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,51200,6.616230010986328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,65536,6.375362820095486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,16384,0.43799556626213926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,65536,8.71502939860026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,16384,1.5784764819675023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,12288,1.232303089565701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,51200,4.910057915581597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,12288,0.3336933453877767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,16384,1.4861475626627605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,12288,0.9468569225735135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,12288,1.1983724170260959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,10240,0.32571289274427623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,10240,0.8289893468221029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,10240,0.7898337576124402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,8192,0.23188177744547525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,8192,0.6521013577779134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,10240,1.143580436706543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,8192,0.7104533513387045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,7168,0.20664000511169434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,8192,0.6772178014119467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,6144,0.4514933162265354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,7168,0.6231297916836208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,7168,0.5953653123643663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,6144,0.17721066210005018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,6144,0.5291440221998427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,5120,0.4557013511657715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,6144,0.4831546677483453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,5120,0.15435377756754556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,5120,0.49537865320841473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,4096,0.3707724412282308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,7168,0.5892302195231119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,5120,0.4024888939327664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,4096,0.3791617817348904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,4096,0.1258444388707479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,3584,0.32532087961832684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,4096,0.3264942169189453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,3584,0.11537778377532959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,3072,0.22357334030999076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,3584,0.30460622575547963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,3072,0.10252444611655341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,3584,0.28817688094245064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,2560,0.19512089093526205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,3072,0.3028586705525716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,3072,0.25505243407355416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,2560,0.22628445095486113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,2048,0.17677688598632812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,2560,0.08836533625920613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,1536,0.119851549466451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,2560,0.2171057727601793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,2048,0.07360711362626818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,2048,0.18888178136613634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,1024,0.08818399906158447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,2048,0.17734132872687447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,1536,0.061576002173953585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,1536,0.1482373343573676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,768,0.07795021931330363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,1024,0.050126221444871694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,1536,0.16112889183892146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,1024,0.11319822735256618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,512,0.054198222027884595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,1024,0.1095857752694024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,768,0.04566133353445265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,768,0.08848355876074897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,256,0.03626755542225308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,768,0.09042133225335015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,512,0.041913777589797974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,128,0.02825688984658983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,512,0.07582577731874254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,256,0.037295109695858426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,256,0.06812533405092028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,64,0.026172444224357605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,256,0.058584888776143394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,16384,128,0.03700622253947788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,16384,32,0.026231999198595684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,128,0.06428088744481404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,128,0.04479555620087517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,64,0.06573333342870076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,16384,32,0.06689155764049955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,16384,512,0.07694222529729207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,65536,1.7520142661200628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,65536,4.105235629611545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,51200,3.20231564839681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,16384,1.052678214179145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,51200,1.3179812961154513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,65536,4.746649000379774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,65536,6.6969714694552955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,16384,1.2330133650037978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,16384,0.376127110587226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,51200,3.6492941114637585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,12288,0.8000719812181262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,16384,1.1011501948038738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,12288,0.2543022102779812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,12288,0.8416559961107043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,10240,0.663040902879503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,12288,0.7999839782714844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,51200,5.5516552395290795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,10240,0.2191928890016344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,10240,0.7714382277594672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,8192,0.17657777998182508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,8192,0.5814879735310873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,8192,0.5774186452229818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,7168,0.4901288880242242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,8192,0.4991404215494792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,7168,0.15825955073038736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,7168,0.4695315361022949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,7168,0.4696604410807292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,6144,0.45481687121921116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,6144,0.13755910926394993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,6144,0.420397334628635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,10240,0.6225146717495388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,5120,0.3327884409162733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,6144,0.36855377091301816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,5120,0.12072443962097168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,4096,0.2653440104590522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,5120,0.437101337644789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,4096,0.09968444373872544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,5120,0.30868532922532826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,3584,0.2192160023583306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,4096,0.2880106767018636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,3584,0.08726400136947632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,4096,0.25425510936313206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,3072,0.20260000228881836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,3584,0.23909065458509657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,3072,0.07761777771843804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,2560,0.16682310899098715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,3584,0.22665866216023764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,3072,0.20639467239379883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,2048,0.14741422070397273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,3072,0.1956862211227417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,2560,0.07856711414125231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,2560,0.17451822757720947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,1536,0.11225688457489014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,2048,0.05758577585220337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,2560,0.16721244653066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,1024,0.07780533366733126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,2048,0.152218672964308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,1536,0.11369600560930039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,1536,0.11689155631595188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,768,0.05515199899673462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,1024,0.04039555456903245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,1024,0.08630400233798557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,1024,0.08978399965498184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,512,0.04184800055291918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,768,0.03723377651638455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,768,0.07080177466074626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,768,0.07253422339757283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,256,0.02879377868440416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,512,0.03405333227581448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,512,0.05904533465703329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,512,0.05962933434380425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,2048,0.14425599575042725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,128,0.022615111536449854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,256,0.05275555451711019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,51200,15.840135362413195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,256,0.047298666503694325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,1536,0.047799110412597656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,64,0.030464887619018555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,128,0.03064089020093282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,12288,32,0.02072355482313368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,128,0.04993155598640442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,12288,128,0.03596444593535529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,64,0.05065244436264038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,12288,32,0.05952178107367622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,12288,256,0.030751999881532457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,65536,1.4727662404378254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,51200,2.942153718736437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,51200,1.0629564921061199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,65536,3.4698925018310547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,16384,0.8683848910861545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,51200,4.151661343044705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,16384,1.0839582019382052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,16384,0.9147208531697592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,51200,3.1923014322916665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,65536,4.031500498453776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,65536,5.228019714355469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,12288,0.6550559997558594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,12288,0.24490311410692003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,12288,0.717465771569146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,12288,0.673682689666748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,10240,0.6164097785949707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,10240,0.18797243965996635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,10240,0.6025990910000271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,10240,0.5332978036668565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,8192,0.4673120180765788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,8192,0.14851289325290257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,8192,0.5961350864834255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,8192,0.4153289000193278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,7168,0.39968265427483457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,7168,0.13157067033979628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,7168,0.36495288213094074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,6144,0.34763820966084796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,7168,0.5549386872185601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,6144,0.11625777350531684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,6144,0.39543644587198895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,5120,0.29209423065185547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,6144,0.31080177095201283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,5120,0.2880764537387424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,5120,0.10183199908998276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,4096,0.2271937794155545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,4096,0.0842311117384169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,3584,0.181807107395596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,5120,0.2643653286827935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,4096,0.231769773695204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,3584,0.07388889127307467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,4096,0.21630756060282388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,3584,0.212647106912401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,3072,0.1613866620593601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,2560,0.1345048877927992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,3584,0.19282044304741752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,3072,0.1756453381644355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,2560,0.059382220109303795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,2560,0.14896178245544434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,2048,0.1067075530687968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,2560,0.14433688587612575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,1536,0.08871022197935317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,2048,0.12166666984558105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,2048,0.12221688694424099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,1536,0.04190488987498813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,1536,0.1004142231411404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,1024,0.06181244717703926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,1536,0.09754577610227798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,1024,0.03559022148450216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,16384,0.28412532806396484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,1024,0.0740106635623508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,768,0.04537777768241035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,51200,65536,20.602323744032116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,3072,0.16746311717563203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,768,0.03315644462903341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,512,0.04734933376312256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,768,0.05960444609324137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,256,0.02964444292916192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,768,0.06319910950130887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,512,0.04442310995525784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,512,0.05065155691570706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,128,0.021298666795094807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,256,0.04008533226119147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,256,0.04529777831501431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,256,0.041217777464124895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,64,0.018031999468803406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,128,0.04253244400024414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,10240,32,0.017835555805100333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,128,0.03124177787039015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,3072,0.06656622224383883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,64,0.04337155487802294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,10240,32,0.044171555174721606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,1024,0.07553866836759779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,2048,0.049350221951802574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,10240,512,0.052389333645502724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,10240,128,0.027291556199391682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,65536,1.2323368920220268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,51200,0.9378532833523221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,51200,2.1105333964029946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,16384,0.9256418016221789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,65536,2.8630907270643444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,16384,0.25984621047973633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,12288,0.5544657707214355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,16384,0.7103022469414605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,16384,0.9724320305718316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,12288,0.1951528920067681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,51200,3.405432807074653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,10240,0.4421999984317356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,51200,2.564459482828776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,65536,3.2206933763292103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,10240,0.18665599822998047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,65536,4.342144012451172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,12288,0.543800883822971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,8192,0.39591288566589355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,10240,0.49385955598619247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,8192,0.12577777438693577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,10240,0.42772356669108075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,7168,0.3209902180565728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,8192,0.43840445412529844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,7168,0.11231022410922581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,6144,0.24934310383266875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,7168,0.3936897648705377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,8192,0.43616533279418945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,7168,0.3074471155802409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,6144,0.09911733203464085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,5120,0.21532266669803193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,12288,0.5783084233601888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,5120,0.08529866404003567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,6144,0.25876977708604604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,4096,0.17343110508388945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,5120,0.23640799522399902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,3584,0.14806044101715088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,4096,0.07101600037680732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,5120,0.22363111707899305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,3072,0.12726755936940512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,4096,0.17908444669511583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,3584,0.06366222434573703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,3584,0.16228355301751032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,2560,0.1208604441748725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,3072,0.05650666687223646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,3584,0.18079288800557455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,3072,0.14653155538770887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,3072,0.13996622297498915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,2048,0.09037688705656265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,2560,0.12128267023298477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,6144,0.3225439919365777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,2560,0.11953866481781006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,2048,0.04230755567550659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,2048,0.10034844610426162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,1024,0.04741155438952976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,2048,0.10207288795047337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,1536,0.036052445570627846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,4096,0.18344444698757598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,1536,0.07962844769159953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,768,0.04096622268358866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,1536,0.08188088734944661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,1024,0.060737775431738965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,512,0.026978666583697002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,1024,0.06391111347410414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,768,0.027406222290462915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,768,0.04878133204248217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,256,0.019520888725916546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,768,0.05494399865468343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,512,0.02456444501876831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,512,0.04076000054677328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,2560,0.05087111062473721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,512,0.04462755719820658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,128,0.015237333046065437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,256,0.02277155551645491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,64,0.013810666898886362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,256,0.03537688983811273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,1536,0.07647111018498738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,8192,32,0.013774221969975365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,128,0.022631999519136217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,128,0.03412177827623155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,64,0.03455288873778449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,32,0.03458311160405477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,8192,1024,0.030039111773173015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,8192,256,0.03626133335961236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,8192,128,0.027088888817363318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,65536,1.097008917066786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,16384,0.8753386603461372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,51200,0.8180453512403699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,51200,2.321410708957248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,65536,2.6796302795410156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,16384,0.23606933487786186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,12288,0.5618231031629775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,16384,0.87426667743259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,16384,0.6412142117818197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,51200,3.286681705050998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,65536,2.888121710883247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,12288,0.5463840166727701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,65536,3.934053209092882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,51200,2.181990305582682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,12288,0.16978666517469618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,10240,0.3940888775719537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,10240,0.13964000013139513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,12288,0.46268002192179364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,8192,0.3251689010196262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,10240,0.45862579345703125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,7168,0.29436445236206055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,10240,0.38587111896938747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,8192,0.4072648949093289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,8192,0.30732178688049316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,7168,0.09830488761266072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,6144,0.22790845235188803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,7168,0.29985777537027997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,6144,0.09001244439019097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,5120,0.2015297810236613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,6144,0.2550853358374702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,4096,0.1459368864695231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,6144,0.2346408896976047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,5120,0.22651643223232695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,5120,0.19723999500274658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,4096,0.06370044416851468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,4096,0.16564888424343532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,4096,0.16745332876841226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,8192,0.11675911479526097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,3584,0.1447519991132948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,3584,0.056812445322672524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,3072,0.12027555041842991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,3584,0.14300533135732016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,2560,0.090737779935201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,3072,0.05122133427196079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,3072,0.1264506710900201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,3072,0.12562044461568198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,7168,0.2999697791205512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,2560,0.10859911309348212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,2048,0.0794995559586419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,2560,0.045055111249287925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,2560,0.11053333017561172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,1536,0.0658737752172682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,2048,0.03831377625465393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,2048,0.0896008875634935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,3584,0.13712533315022787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,1024,0.046219554212358266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,2048,0.09090755383173625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,1536,0.07115999857584636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,768,0.03348533312479655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,1024,0.02697333362367418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,1024,0.05432000093989902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,1024,0.05769689215554131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,512,0.025458667013380263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,768,0.04401866594950358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,768,0.02446222139729394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,768,0.04857244425349765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,256,0.01779733267095354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,512,0.035523557000690036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,512,0.04099999864896139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,128,0.0144195556640625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,256,0.01994400057527754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,256,0.030913776821560327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,256,0.032103111346562706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,128,0.01953866746690538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,64,0.013045333325862885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,1536,0.03199999862247043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,128,0.029823111163245306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,128,0.024752888414594863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,7168,1536,0.07468622260623507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,64,0.029652443197038438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,7168,32,0.02962133288383484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,512,0.021536888347731695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,7168,5120,0.07756266991297404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,7168,32,0.013389333254761167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,65536,0.9057804743448893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,51200,1.7352541817559137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,65536,2.0574052598741317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,51200,0.6765102280510796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,16384,0.7636106808980306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,16384,0.21342133151160347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,12288,0.42915911144680446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,16384,0.6311448944939507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,16384,0.5747671127319336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,12288,0.1465511057111952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,65536,3.2861866421169705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,65536,2.558126237657335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,12288,0.5480648676554362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,10240,0.33689244588216144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,51200,2.8831111060248484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,51200,2.0154221852620444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,12288,0.4315902127159967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,10240,0.1314444409476386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,8192,0.29696088367038304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,8192,0.11422222190433079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,7168,0.2505066659715441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,8192,0.372162659962972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,10240,0.3481608761681451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,6144,0.19894933700561523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,7168,0.09078933133019342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,8192,0.29351377487182617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,7168,0.2616648938920763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,5120,0.1710453298356798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,7168,0.2761502265930176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,6144,0.08287022511164348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,6144,0.22932799657185873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,4096,0.13600177235073513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,5120,0.06818311081992255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,5120,0.1818337837855021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,3584,0.12072000238630508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,4096,0.05960710843404134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,5120,0.18727733029259575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,4096,0.15005244149102107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,3072,0.09732888804541694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,3584,0.05249155561129252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,4096,0.14914045068952772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,2560,0.07337511248058744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,10240,0.36344001028272843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,3072,0.045575999551349215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,3072,0.11259110768636067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,2048,0.0651128888130188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,3072,0.11790400081210667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,2560,0.04018400112787882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,2560,0.09574666950437759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,2048,0.03443288803100586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,2560,0.09971111350589329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,1536,0.05657511287265354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,2048,0.07951911290486653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,1024,0.033794667985704206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,2048,0.08440799845589532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,1536,0.028578665521409776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,6144,0.2142337825563219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,768,0.02716355522473653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,1536,0.06263111035029094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,1536,0.06846488846672906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,1024,0.024275556206703186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,1024,0.04775111211670769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,512,0.019401777121755812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,1024,0.05344177616967095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,768,0.02181244393189748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,768,0.03879466652870178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,3584,0.12681066989898682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,256,0.01519466605451372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,768,0.0448053346739875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,512,0.0196079992585712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,512,0.031139555904600356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,128,0.01202311118443807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,512,0.03682755430539449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,256,0.02713511056370205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,64,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,256,0.028402666250864666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,128,0.018290667070282828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,128,0.025778666138648987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,128,0.021877333521842957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,64,0.025987555583318073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,6144,32,0.0258240004380544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,6144,3584,0.1362480057610406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,6144,256,0.017810665898852877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,65536,0.7202000088161893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,6144,32,0.011038222246699862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,65536,1.8753857082790797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,16384,0.41205067104763454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,51200,1.559931543138292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,16384,0.5251759952969021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,65536,2.795017878214518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,16384,0.16617243819766575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,16384,0.4981688923305935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,12288,0.3451111051771376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,12288,0.1261519988377889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,51200,2.1806640625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,10240,0.2568133407168918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,51200,0.5651991102430556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,51200,1.6453644434611003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,12288,0.3901253276401096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,12288,0.3697644339667426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,10240,0.11008266607920329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,8192,0.2463395595550537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,10240,0.31427110566033256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,8192,0.0877493355009291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,7168,0.21692978011237252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,8192,0.27833599514431423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,10240,0.3113439877827962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,8192,0.24401156107584634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,7168,0.07867022355397542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,6144,0.16798844602372912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,7168,0.21186044481065539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,65536,2.192437277899848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,5120,0.14281955030229357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,6144,0.06778044170803495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,7168,0.214227557182312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,4096,0.117850661277771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,5120,0.05937066343095568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,6144,0.21180266804165312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,6144,0.1921751101811727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,3584,0.09419288900163438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,4096,0.0537484453784095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,5120,0.1579937802420722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,4096,0.12408355871836345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,3072,0.08989066547817653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,3584,0.044414222240448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,4096,0.13011466132269964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,3584,0.10947200324800278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,2560,0.06834755341211955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,3072,0.040714667903052434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,3584,0.11694133281707764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,3072,0.095396445857154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,2048,0.059747556845347084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,2560,0.03462222218513489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,3072,0.1023448904355367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,2560,0.08197955290476482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,2048,0.02996266550487942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,2560,0.08818311161465114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,2048,0.06744888756010267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,1024,0.031629333893458046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,2048,0.07526310947206286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,1536,0.025447110335032146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,1536,0.05404000149832832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,768,0.024194666081004675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,1536,0.06049599912431505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,1024,0.0214008887608846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,1024,0.0410426656405131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,512,0.017294221454196505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,1024,0.04740177922778659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,5120,0.17578489250606963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,768,0.019162666466501024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,768,0.033655110332700945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,256,0.012748444245921241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,768,0.03992088966899448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,512,0.01790844400723775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,128,0.013734222286277346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,512,0.02718399961789449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,256,0.016248888439602323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,512,0.03350933392842611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,64,0.012720888687504662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,256,0.023289778166347083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,5120,128,0.016360888878504436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,256,0.02502311103873783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,128,0.02236000034544203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,5120,128,0.0199946661790212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,64,0.022327999273935955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,1536,0.041835556427637734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,5120,32,0.022237333986494277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,65536,0.5814631250169542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,5120,32,0.011687111523416309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,65536,1.3295759624905057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,51200,1.1588968700832791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,16384,0.36360979080200195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,51200,0.46981064478556317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,16384,0.4530160162183974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,51200,1.7090106540256078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,65536,2.00815561082628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,16384,0.14150488376617432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,12288,0.2714657783508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,65536,1.8859049479166667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,12288,0.1061591108640035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,10240,0.22679556740654838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,16384,0.4551839828491211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,12288,0.29812799559699166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,10240,0.09283822112613255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,12288,0.3173164526621501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,10240,0.25889600647820366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,51200,1.4002275466918945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,8192,0.07286755243937175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,7168,0.1697982284757826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,8192,0.20390578111012778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,8192,0.20822044213612875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,10240,0.2650008996327718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,6144,0.14075377252366808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,7168,0.06570933262507121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,7168,0.1803333361943563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,5120,0.10263822476069133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,6144,0.05864266554514567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,7168,0.19584622648027208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,6144,0.15324444240993923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,5120,0.05479733480347527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,6144,0.15891733434465197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,5120,0.12945600350697836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,5120,0.1352657741970486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,3584,0.08223022354973687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,4096,0.10520799954732259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,4096,0.04319377740224203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,3072,0.07128177748786078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,3584,0.03756977783309089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,4096,0.11265867286258274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,8192,0.16426577832963732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,3584,0.09258488814036052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,2560,0.05613155497445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,3072,0.03344444433848063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,3584,0.09957155254152085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,3072,0.08294488986333211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,2048,0.04594044552909004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,2560,0.02938844429122077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,1536,0.034062223301993474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,3072,0.08821600013309056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,2560,0.06872889068391588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,2048,0.02534399926662445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,2560,0.07623200284110175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,1024,0.024477332830429077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,2048,0.05695466862784492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,1536,0.021887999441888597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,2048,0.0646159980032179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,768,0.018326222896575928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,1536,0.045695109499825366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,1024,0.018586667047606576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,1536,0.05324088864856296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,4096,0.08941333161460029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,512,0.014102222190962898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,1024,0.034480889638264976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,768,0.016872000363137987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,256,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,768,0.028576887316173975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,512,0.01552177800072564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,768,0.03506311111980014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,128,0.009471110999584198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,512,0.022701332966486614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,256,0.014452444182501899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,512,0.027860442797342937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,64,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,256,0.01963644391960568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,4096,32,0.008603555460770925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,256,0.022318222456508215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,4096,128,0.014465777410401238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,128,0.018862222631772358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,128,0.017623111605644226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,64,0.018579555882347953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,4096,32,0.01864177816443973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,65536,0.5397093560960557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,4096,1024,0.04145244426197476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,51200,1.0485821829901802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,65536,1.2802222569783528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,16384,0.35781333181593156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,51200,0.4311573240492079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,16384,0.12509777810838488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,51200,1.4847884707980687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,12288,0.2580080032348633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,16384,0.3791200055016412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,51200,1.2534462610880535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,65536,1.7844098409016926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,16384,0.38111289342244464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,65536,1.604847060309516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,12288,0.0982017781999376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,12288,0.2685040103064643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,10240,0.23279288080003527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,10240,0.09106933408313328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,10240,0.2327200041876899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,12288,0.3024142318301731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,8192,0.18014666769239637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,10240,0.24454755253261992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,7168,0.16503911548190647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,8192,0.20892533991071913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,6144,0.12616356213887533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,7168,0.06003377834955851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,7168,0.16458133856455484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,6144,0.05305777655707466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,5120,0.10613599750730728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,7168,0.1672391096750895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,6144,0.14117333624098036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,4096,0.09124888976414998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,6144,0.1516231165991889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,5120,0.044700443744659424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,5120,0.11620622211032444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,4096,0.03660800059636434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,3584,0.0819644464386834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,8192,0.1834906604554918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,4096,0.09496444463729858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,4096,0.10673777924643622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,3072,0.06931733422809176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,3584,0.033746666378445096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,8192,0.06795466608471341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,3584,0.08303378025690715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,2560,0.053351110882229276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,3584,0.09115999937057495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,3072,0.02996088729964362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,3072,0.07282488875918917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,2048,0.04383466641108195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,3072,0.08138222164577909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,2560,0.02665333284272088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,2560,0.061390221118927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,2048,0.02306933369901445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,2560,0.07065511412090726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,1024,0.0211217784219318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,2048,0.051664888858795166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,2048,0.059980445437961154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,1536,0.019472888774342008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,1536,0.04124266571468777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,768,0.018401778406567044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,1536,0.047919111119376294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,1024,0.016755556066830952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,1024,0.03077510992685954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,512,0.014640889234013028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,5120,0.12452533509996201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,1024,0.03729866610633002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,768,0.01574844453069899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,256,0.011789333489206104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,768,0.025566221939192876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,512,0.014469333820872836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,768,0.031275557147132024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,128,0.00945955514907837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,512,0.02056533263789283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,256,0.013585777746306526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,512,0.02677333354949951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,256,0.017847110827763874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,64,0.007823111282454597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3584,128,0.013561777770519257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,32,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,128,0.017462222112549674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,256,0.022226666410764057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3584,128,0.017080888152122498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,64,0.016924444172117446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3584,1536,0.03419555558098687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3584,32,0.016885333591037326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,65536,0.459353764851888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,51200,0.3650035593244765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,51200,0.7288968827989367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,16384,0.27280622058444554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,65536,1.1845742331610787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,16384,0.11837155289120144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,16384,0.3226115438673231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,12288,0.2401155630747477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,51200,1.2463493347167969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,16384,0.35249866379631895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,65536,1.5153413348727758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,51200,1.1436595916748047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,12288,0.1090666651725769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,10240,0.1817457808388604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,12288,0.2643057770199246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,65536,1.4548755221896703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,8192,0.16530666086408827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,10240,0.0838622252146403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,12288,0.2530053456624349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,10240,0.205805328157213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,7168,0.11982222398122151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,8192,0.06637955374187894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,10240,0.2352693345811632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,6144,0.09981866677602132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,8192,0.18099555704328749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,7168,0.053371555275387235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,8192,0.17614577876196968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,7168,0.14780622058444554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,5120,0.0909493300649855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,6144,0.05101866523424784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,7168,0.15308977497948542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,6144,0.13292977544996473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,4096,0.06907555792066786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,5120,0.039452443520228066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,6144,0.1525191068649292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,3584,0.061884442965189614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,4096,0.0340008901225196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,5120,0.10460266802046035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,4096,0.08374310864342584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,5120,0.11595200167761908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,3072,0.051426665650473706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,3584,0.030015110969543457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,4096,0.09420977698432074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,2560,0.041322665082083807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,3584,0.0751475559340583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,3584,0.08323378033108182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,3072,0.06441422303517659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,2048,0.03518844313091702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,3072,0.07337688737445407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,2560,0.02420177724626329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,2560,0.05589422252443102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,1536,0.027441778116756018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,2560,0.06531377633412679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,2048,0.02107644412252638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,1024,0.01831022236082289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,2048,0.046557333734300405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,2048,0.05394488904211256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,1536,0.037335998482174344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,1536,0.018415111634466384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,768,0.015191111299726697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,1536,0.04483911063936022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,1024,0.015928000211715698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,1024,0.02789333462715149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,512,0.011296889020336999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,768,0.01482488877243466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,768,0.023353776997990076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,256,0.008570666942331526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,512,0.013627555635240344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,768,0.028856890069113836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,128,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,512,0.01865511139233907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,512,0.02402666707833608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,256,0.012780444489585029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,256,0.016210666961140104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,3072,0.02683199942111969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,3072,128,0.012777778009573618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,32,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,128,0.01605244477589925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,64,0.015121777852376303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,32,0.01516888870133294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,1024,0.034168889125188194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,65536,0.4158951176537408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,65536,0.8627173105875651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,3072,64,0.006559111177921295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,3072,256,0.021000888612535264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,3072,128,0.015505777464972602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,65536,1.1961253484090169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,16384,0.2493751049041748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,51200,0.7271662288241916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,51200,0.3304684427049425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,65536,1.2546488444010417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,12288,0.19332533412509492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,16384,0.27048712306552464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,16384,0.09999377859963311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,10240,0.16053866015540227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,12288,0.08257333437601726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,16384,0.3071102301279704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,12288,0.20790221956041124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,51200,1.036963568793403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,51200,0.9579955206976997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,10240,0.18711288770039877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,8192,0.13412533866034612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,10240,0.07125155793295966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,12288,0.24021244049072266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,7168,0.11372710598839654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,8192,0.05858400132921007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,10240,0.1881706714630127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,7168,0.047882666190465294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,8192,0.1416915522681342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,6144,0.10142666763729519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,8192,0.1660782231224908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,5120,0.07951911290486653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,6144,0.04111288984616598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,7168,0.13835822211371526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,4096,0.06083910995059543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,6144,0.11244889100392659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,5120,0.03490488727887472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,5120,0.09125689003202651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,6144,0.1194559997982449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,3584,0.05627110931608412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,5120,0.10257066620720758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,4096,0.07330399751663208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,3072,0.04886933167775472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,3584,0.02643644478585985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,3584,0.0654239985677931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,4096,0.08454933431413438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,2560,0.04047200083732605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,3584,0.0755902197625902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,3072,0.02405422263675266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,3072,0.05646222167544895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,2048,0.034091555409961276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,3072,0.06604355573654175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,2560,0.021446221404605444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,2560,0.0491502218776279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,1536,0.026550221774313185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,2560,0.05814044343100654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,2048,0.01899377836121453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,2048,0.04087911049524943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,7168,0.12627822822994655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,1024,0.016121778223249648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,2048,0.049182222949133984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,1536,0.01681777834892273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,768,0.013002666334311167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,1024,0.01447111037042406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,1536,0.040680001179377236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,1024,0.024846222665574815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,4096,0.030431999100579157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,1024,0.030624889665179785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,512,0.010784000158309937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,768,0.013731555806265937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,768,0.02070399953259362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,256,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,512,0.012732444538010491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,512,0.017131555411550734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,512,0.022063111265500385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,256,0.014434667097197639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,256,0.012081777883900536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,128,0.006654222392373615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,64,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,256,0.018215111560291715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2560,32,0.0079644446571668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2560,128,0.012091555529170565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,128,0.013775111072593264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,128,0.015238222148683337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,64,0.013548444542619916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,32,0.013570666313171387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2560,1536,0.032871110571755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,65536,0.6194373236762153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,65536,0.3332480059729682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,51200,0.533514658610026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2560,768,0.02642400066057841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,16384,0.18019021881951225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,51200,0.26528533299763996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,16384,0.08510400189293756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,65536,0.9950906965467664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,12288,0.13441867298550075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,51200,0.7711368666754829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,16384,0.27575911415947807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,16384,0.2569422192043728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,65536,1.0223208533393013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,10240,0.1259253289964464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,12288,0.07227466503779094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,8192,0.09452178080876668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,12288,0.17683645089467367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,10240,0.061894224749671094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,12288,0.20435555775960287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,51200,0.7926817999945747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,10240,0.15147022406260172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,7168,0.08549155791600545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,8192,0.04508799976772732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,10240,0.16598133246103922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,6144,0.07465689049826728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,8192,0.12418311172061497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,7168,0.0406640006436242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,5120,0.05903644694222344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,8192,0.13981066809760198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,7168,0.10680355628331502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,6144,0.03444000085194906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,4096,0.04683466752370199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,7168,0.12595289283328587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,6144,0.09082133240169948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,5120,0.030202666918436687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,3584,0.040576001008351646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,6144,0.10673599772983128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,4096,0.025751999682850305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,5120,0.09188355339898004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,3072,0.03368622064590454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,4096,0.062487112151251904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,3584,0.023418666587935552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,4096,0.07353866762585111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,2560,0.028483556376563177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,3584,0.055664002895355225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,3072,0.020945777495702107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,3584,0.06620177957746717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,2048,0.023447111248970032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,3072,0.04865688747829861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,2560,0.019279999865425956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,1536,0.019023999571800232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,3072,0.0586400032043457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,2560,0.041909333732393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,2048,0.01720711092154185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,2560,0.05207644568549263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,2048,0.03547822104560004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,1024,0.01388177772363027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,2048,0.0441599984963735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,1536,0.015368888775507608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,1536,0.028558221128251817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,768,0.01089599976936976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,1536,0.0355333321624332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,1024,0.02188088827662998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,1024,0.013241777817408243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,512,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,1024,0.02738933265209198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,768,0.012688889271683164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,768,0.018066666192478605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,256,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,768,0.023669333921538457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,5120,0.07713688744439019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,512,0.01178311142656538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,512,0.015207111835479736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,128,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,256,0.011034666664070554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,512,0.020048000746303134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,64,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,256,0.012744888663291931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,2048,128,0.011392888923486074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,2048,32,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,256,0.016620443926917184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,128,0.012180444267061023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,2048,128,0.013401777380042605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,64,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,2048,32,0.011984888878133563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,65536,0.2776462237040202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,65536,0.5275031195746528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,51200,0.3605831199222141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,16384,0.14267288313971624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,51200,0.21989333629608154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,65536,0.8148595492045084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,16384,0.2020346721013387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,51200,0.6322071287367079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,12288,0.10649066501193577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,16384,0.07423822085062663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,65536,0.8522702323065864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,51200,0.6665075620015463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,16384,0.23130578464931914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,10240,0.09746488597657944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,12288,0.06297066476609972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,8192,0.07986844248241849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,12288,0.1508328914642334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,10240,0.055665777789221875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,12288,0.1764613389968872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,7168,0.0609804458088345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,10240,0.12977066304948595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,8192,0.042414221498701304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,10240,0.14688532882266575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,8192,0.10082222355736627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,6144,0.056701335642072886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,5120,0.04620000057750278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,7168,0.03846844368510776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,7168,0.08956533008151585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,6144,0.03196000059445699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,4096,0.035924444595972695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,6144,0.07659644550747342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,7168,0.10517244206534491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,5120,0.027860442797342937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,6144,0.09112799829906887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,3584,0.029215110672844782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,5120,0.06464533011118571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,4096,0.023571555813153584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,5120,0.07619555791219075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,3072,0.0258879992696974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,4096,0.052822222312291466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,3584,0.02122755514250861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,2560,0.0232933329211341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,4096,0.06418133444256253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,3584,0.04730755421850416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,3072,0.018718222777048748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,3584,0.05795733133951823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,2048,0.01811911165714264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,3072,0.04189688960711161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,2560,0.017831999394628737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,1536,0.014721777704026965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,3072,0.051613334152433604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,2560,0.03625955515437656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,2048,0.015832889411184523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,2560,0.04432355695300632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,1024,0.01105955574247572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,2048,0.030352890491485596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,1536,0.013798221945762634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,2048,0.03745422098371718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,1536,0.024728000164031982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,1536,0.030586666531032983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,1024,0.012061333490742577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,8192,0.12139556143018936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,512,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,1024,0.024297777149412367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,768,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,768,0.015468445089128284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,256,0.0053484443989064955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,768,0.020937777227825586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,512,0.012823999755912356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,512,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,128,0.005026666654480828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,512,0.018927999668651156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,256,0.01034311122364468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,256,0.011057777537239922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,64,0.00452888881166776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,256,0.01571911076704661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1536,128,0.010395555860466426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,32,0.004556444370084339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,128,0.010363555616802638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1536,128,0.012421333127551608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,64,0.010000000397364298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,32,0.010067555639478896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,65536,0.368392891354031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1536,768,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,65536,0.22462222311231825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,51200,0.29156266318427193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1536,1024,0.018668444620238412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,16384,0.09718577729331122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,65536,0.6072070863511827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,51200,0.1786115566889445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,12288,0.07555555635028414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,16384,0.0587315559387207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,16384,0.15534488360087076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,51200,0.5181244214375814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,65536,0.7017048729790581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,10240,0.0572906666331821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,16384,0.1998035510381063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,12288,0.047135998805363975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,51200,0.5519164403279623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,8192,0.04956266615125868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,12288,0.11785333686404759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,10240,0.042464001311196216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,10240,0.10189244482252334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,12288,0.1436986658308241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,7168,0.043262223402659096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,8192,0.03102311160829332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,10240,0.1256399949391683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,8192,0.08211021953158908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,6144,0.03732000125779046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,7168,0.02714577813943227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,5120,0.028484443823496502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,8192,0.1002311110496521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,6144,0.023709333605236475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,7168,0.08885955810546875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,4096,0.023222222924232483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,6144,0.06083199712965223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,5120,0.020675554871559143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,6144,0.07657511366738214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,3584,0.02155911094612545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,5120,0.05150399936570061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,4096,0.0177528891298506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,5120,0.06525244315465291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,4096,0.043169776598612465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,3584,0.01631733278433482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,4096,0.05481777919663323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,2560,0.016113777955373127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,3584,0.03826489051183065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,3584,0.049060444037119545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,3072,0.034028444025251604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,2048,0.01367288910680347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,3072,0.04385155439376831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,2560,0.01404266721672482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,2560,0.029172443681293066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,1536,0.010753778119881948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,2560,0.03716444306903415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,2048,0.012843555874294706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,2048,0.024879111184014216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,1024,0.008763555851247575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,2048,0.03239999877081977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,1536,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,7168,0.07237422466278076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,768,0.007247111035717859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,1536,0.026094221406512793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,1024,0.010741333166758219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,1024,0.015508444772826301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,512,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,1024,0.021163556310865615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,768,0.010392888552612728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,768,0.01310755560795466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,3072,0.0192275560564465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,256,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,768,0.01854222185081906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,512,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,512,0.011381333072980246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,256,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,3072,0.015125332607163323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,512,0.015789333317014907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,256,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,64,0.003504000190231535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,256,0.013777777552604675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,1024,128,0.009328888522254096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,128,0.00865955568022198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,1024,128,0.010766222245163388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,64,0.008352888955010308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,32,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,65536,0.31815732849968803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,1024,1536,0.020291555258962844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,51200,0.23526043362087676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,65536,0.2118177811304728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,16384,0.07215822405285306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,65536,0.5237529012891982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,51200,0.16993244489034018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,128,0.003867555409669876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,51200,0.44287821981641984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,16384,0.1348960002263387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,12288,0.0588666664229499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,65536,0.6967288653055826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,1024,32,0.0035573335157500375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,16384,0.058443552917904325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,51200,0.5480453173319498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,10240,0.048222220606274076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,16384,0.18339822027418348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,12288,0.04639822244644165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,8192,0.037279999918407865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,12288,0.10315022203657363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,10240,0.04112977782885233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,7168,0.033313777711656355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,12288,0.14428622192806667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,10240,0.08651910887824164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,8192,0.02968177861637539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,10240,0.12053778436448838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,6144,0.02701955537001292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,8192,0.06999999947018094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,7168,0.026354667213228013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,8192,0.09848711225721572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,5120,0.02345422241422865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,7168,0.06167733007007175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,6144,0.023200000325838726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,4096,0.018315555320845712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,7168,0.08727644549475776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,6144,0.05305066704750061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,5120,0.02036355601416694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,6144,0.07521155807707045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,3584,0.01680799987581041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,5120,0.04536799920929802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,4096,0.01770400007565816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,5120,0.06405066781573825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,3072,0.01461866663561927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,4096,0.037435554795795016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,3584,0.016529776983790927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,2560,0.012950222525331708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,4096,0.05331200030114916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,3584,0.033681776788499616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,3072,0.015170666906568738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,3584,0.04782399866316053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,2048,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,3072,0.029324445459577773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,3072,0.04180888997183906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,1536,0.00903555585278405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,2560,0.02570666703912947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,2048,0.013080889152155982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,2560,0.036460445986853704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,1024,0.00700711127784517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,2048,0.021721776988771226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,2048,0.03108000093036228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,1536,0.01180355581972334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,768,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,1536,0.01792711185084449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,1024,0.01073688848151101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,512,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,1024,0.013835555149449242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,768,0.010323555933104621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,1024,0.02071822186311086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,256,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,768,0.012106666962305704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,768,0.018447111050287884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,512,0.009711110757456886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,512,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,128,0.003504000190231535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,512,0.015496888094478183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,256,0.009304000271691216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,256,0.00834222220712238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,64,0.003168888803985384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,768,32,0.003154666680428717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,256,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,128,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,128,0.010689777632554373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,64,0.007366221812036302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,65536,0.21342399385240343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,2560,0.014126222994592456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,51200,0.16264443927341038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,65536,0.16854311360253227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,65536,0.4384026792314317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,768,1536,0.025774222281244066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,16384,0.05704355239868164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,51200,0.3428942097557916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,51200,0.13464978006150988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,12288,0.04125244418780009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,65536,0.5555466545952691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,16384,0.04669777883423699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,16384,0.11463377210828994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,10240,0.03376266691419814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,51200,0.47182665930853945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,16384,0.14847377936045328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,12288,0.03772355450524224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,12288,0.08706488874223496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,8192,0.026104888982243005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,12288,0.1152320040596856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,10240,0.03341777788268195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,7168,0.022475555539131165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,10240,0.07322844531801012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,768,32,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,8192,0.023423999547958374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,8192,0.059241778320736356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,8192,0.08025778002209134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,7168,0.020427554845809937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,7168,0.052183998955620654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,7168,0.07013866636488172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,5120,0.016018667154841952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,6144,0.018244443668259513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,6144,0.04528888728883532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,6144,0.06080355246861776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,5120,0.016233777006467182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,4096,0.013392888837390475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,5120,0.038534220721986555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,5120,0.052433778842290245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,3584,0.012108444339699216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,4096,0.014152889450391134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,768,128,0.009311999711725447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,4096,0.03179555469089084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,3072,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,4096,0.04375288883845011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,3584,0.013224889006879596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,3584,0.028811554114023846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,2560,0.009088888764381409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,3584,0.03887288769086202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,10240,0.10112266408072577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,3072,0.025501334004931982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,3072,0.012470222181744046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,6144,0.01904088921017117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,2048,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,3072,0.034684442811542086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,1536,0.006688888702127669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,2560,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,2048,0.011397333608733283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,2560,0.022055110997623865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,1024,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,2048,0.01887999971707662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,2560,0.03016888764169481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,1536,0.010390222072601318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,2048,0.02601066728432973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,1536,0.015505777464972602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,768,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,1024,0.009759999811649323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,1536,0.021792888641357422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,512,0.00388355553150177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,1024,0.012063999970753988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,768,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,256,0.003220444545149803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,1024,0.01789777808719211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,768,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,512,0.008991999758614434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,768,0.015804444750150044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,512,0.009328000247478485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,256,0.008375110725561777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,512,0.013735111388895245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,64,0.0028373334142896864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,256,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,32,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,512,128,0.008386666576067606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,256,0.012074666718641916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,128,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,512,128,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,64,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,65536,0.14330222871568468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,512,32,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,51200,0.11220710807376438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,16384,0.03528533379236857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,65536,0.16568888558281794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,51200,0.13269510534074572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,65536,0.3555404345194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,12288,0.02604266670015123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,51200,0.27989421950446236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,16384,0.09442044628991021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,16384,0.04625688989957174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,10240,0.022056000100241765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,16384,0.14454400539398193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,65536,0.5353031158447266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,51200,0.4213715659247504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,12288,0.03772622346878052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,8192,0.01608444419172075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,512,128,0.0028782222006056043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,12288,0.07125600179036458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,10240,0.03312355611059401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,12288,0.11175999376508926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,8192,0.02300888962215847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,10240,0.059614221254984535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,6144,0.01144088887506061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,10240,0.09489422374301487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,8192,0.0484808882077535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,7168,0.020629333125220407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,8192,0.07759910821914673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,5120,0.010053333308961656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,6144,0.01852888862291972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,7168,0.06774399677912395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,6144,0.03760088814629449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,4096,0.008608000146018134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,6144,0.05925955375035604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,5120,0.01664266652531094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,5120,0.03208177619510227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,3584,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,5120,0.050260444482167564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,4096,0.014136888914638095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,4096,0.026354667213228013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,3072,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,4096,0.041814221276177295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,3584,0.013199110825856527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,3584,0.023623999622133043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,2560,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,3584,0.03717511230044895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,3072,0.021065778202480737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,7168,0.013769778112570444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,2048,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,3072,0.03353955679469638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,2560,0.018529777725537617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,2560,0.01186044431395001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,1536,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,2560,0.029567999972237483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,2048,0.011532444092962476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,7168,0.04285511043336657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,2048,0.015853333804342482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,1024,0.003842666745185852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,2048,0.025062221619817946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,1536,0.010536889235178629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,1536,0.013399110900031196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,768,0.003524444583389494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,1024,0.009666666388511658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,1536,0.020966221888860066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,1024,0.010687111152542962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,512,0.0031715554909573663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,1024,0.0169831116994222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,768,0.009298666483826106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,768,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,256,0.0027502222607533136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,768,0.015488889482286243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,128,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,512,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,512,0.013408000270525614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,256,0.008341333104504479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,64,0.0022275555464956495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,256,0.006303999986913469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,3072,0.012425777812798819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,256,32,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,128,0.008299555215570662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,256,0.011691555380821228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,128,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,65536,0.09412178066041733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,256,128,0.00902844468752543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,64,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,51200,0.07637155718273587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,65536,0.16596710681915283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,16384,0.029483556747436523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,51200,0.13262666596306696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,51200,0.24769955211215547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,16384,0.04751377635531955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,16384,0.08294488986333211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,12288,0.02238133384121789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,51200,0.4151475429534912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,16384,0.14269867208268908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,256,512,0.008939555949634975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,10240,0.01608088943693373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,12288,0.038479109605153404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,12288,0.0632551113764445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,8192,0.012961777547995249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,12288,0.1104888916015625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,10240,0.05298044284184774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,10240,0.09379911422729492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,256,32,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,7168,0.011776000261306763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,8192,0.023932443724738225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,8192,0.04342488778962029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,6144,0.010381333529949188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,7168,0.020968889196713764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,8192,0.07630755503972371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,65536,0.31523556179470485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,7168,0.038140445947647095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,7168,0.06683822472890218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,6144,0.018933332628673978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,6144,0.033575998412238225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,4096,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,65536,0.5278479788038466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,5120,0.016891555653678048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,5120,0.028647998968760174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,3584,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,4096,0.01479466590616438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,4096,0.023753777146339417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,4096,0.04032977753215366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,3584,0.021408889028761122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,3584,0.013843555417325763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,10240,0.03410044312477112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,3584,0.03668444355328878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,2560,0.005315555466545953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,3072,0.01313244468635983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,3072,0.01904088921017117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,3072,0.032602667808532715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,2048,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,2560,0.012069332930776807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,2560,0.01683022247420417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,5120,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,1536,0.00396799999806616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,2560,0.028792887926101685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,2048,0.011406222151385413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,2048,0.014818666709793938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,2048,0.024713777833514746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,1536,0.012350222302807702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,1024,0.003499555504984326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,1536,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,5120,0.04930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,768,0.0032435556252797446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,1536,0.02045866681469811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,1024,0.009678222239017487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,1024,0.009869332942697737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,3072,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,512,0.0028248888750871024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,1024,0.016869333055284288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,768,0.009415999882751042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,256,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,768,0.008369777765538957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,512,0.008794666992293464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,512,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,128,0.0021182222084866627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,512,0.012747555143303342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,256,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,256,0.011409777734014722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,64,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,128,0.008368888662921058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,128,32,0.002159111201763153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,128,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,65536,0.08928000264697605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,64,0.0053119998839166425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,128,32,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,51200,0.0736008882522583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,16384,0.028071999549865723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,65536,0.3144666618771023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,12288,0.021212443709373474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,16384,0.08320177925957574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,51200,0.24742754300435385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,768,0.014704000618722705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,10240,0.017469333277808297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,8192,0.01255022237698237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,7168,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4352,128,256,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,6144,0.00980088859796524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,5120,0.008349333372380998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,12288,0.06292088826497395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,10240,0.05292888813548618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,8192,0.042985777060190834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,3584,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,7168,0.03793511125776503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,3072,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,6144,0.03291822142071194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,5120,0.028219554159376357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,4096,0.02362844513522254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,2048,0.004569777597983678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,3584,0.02128711177243127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,1536,0.0038844446341196695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,3072,0.019036443697081674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,1024,0.003181333343187968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,2560,0.01680355601840549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,768,0.0028693332440323303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,2048,0.014799111419253878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,512,0.002623999905255106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,1536,0.012220444778601328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,256,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,1024,0.009884444375832876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,128,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,768,0.008426666259765625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,64,0.0018959999498393799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,512,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,32,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,256,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,128,0.005335999859703912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,4096,0.0071928890215026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,65536,0.09375733137130737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,51200,0.07454488674799602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,16384,0.026263111167483862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,64,2560,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,12288,0.020040889581044514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,65536,0.31728267669677734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,51200,0.24725156360202363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,10240,0.012113778127564324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,16384,0.0826044413778517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,8192,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,7168,0.009391110804345872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,12288,0.06281599733564588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,6144,0.008344888687133789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,10240,0.05300266544024149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,5120,0.007728888756699032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,8192,0.04287999868392944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,7168,0.037754668129815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,4096,0.007039111521508958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,3584,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,6144,0.03297066688537598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,3072,0.011718221836619906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,5120,0.028239998552534316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,64,0.0052800000541739995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,2560,0.010037333601050908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,3584,0.022095999783939783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,2048,0.004273777620659934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,64,32,0.0052986666560173035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,3072,0.019503111640612285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,6144,0.05819288889567057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,2560,0.017196445001496207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,2048,0.014730667074521383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,1024,0.003168888803985384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,768,0.0028284444577164124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,1536,0.012423111332787408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,512,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,256,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,1024,0.00980177770058314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,768,0.008627555436558193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,128,0.001997333433892992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,512,0.007688889073001013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,256,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,32,0.0018639999131361644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,128,0.0052968888647026485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,64,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,32,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4352,32,4096,0.02372088862790002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,1536,0.003891555385457145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4352,32,64,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,16384,1.8330978817409938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,16384,5.183912913004558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,12288,3.9924719068739147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,16384,6.2200745476616754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,12288,1.3482177522447374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,12288,4.427661471896701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,51200,7.002252366807725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,10240,1.1199804941813152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,10240,3.631535847981771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,10240,3.3563431633843313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,8192,0.959287961324056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,8192,3.0041849348280163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,7168,0.7375413576761881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,8192,2.7477554745144315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,7168,2.419117397732205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,6144,0.6487279997931586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,7168,2.2335092756483292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,6144,2.0056205325656467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,5120,0.5384720166524252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,5120,1.6693003972371419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,5120,1.6882123947143555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,4096,0.4622773594326443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,4096,1.6152764426337347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,6144,2.0479823218451605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,3584,0.4188648859659831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,3584,1.278853310479058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,3584,1.1987253824869792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,3072,0.9045510821872287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,3072,0.418704006406996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,4096,1.3324995040893555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,51200,16.575724283854168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,2560,0.8903155856662326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4352,128,128,0.00829777783817715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,3072,1.0429609086778429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,2048,0.5714497566223145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,2560,0.8359262148539225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,2048,0.3136311107211643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,1536,0.2068871127234565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,1536,0.5308497746785482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,1024,0.33591556549072266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,1536,0.5332275496588813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,1024,0.16489244831932917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,768,0.30038311746385365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,512,0.1957075595855713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,768,0.14741955863104925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,1024,0.43341154522365993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,2048,0.6829217804802789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,768,0.3469724390241835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,256,0.13932622803582087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,512,0.13451911343468562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,128,0.11681155363718669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,512,0.34892090161641437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,64,0.10087111261155869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,256,0.242287105984158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,128,0.11422488424513076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,65536,32,0.10699022478527492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,128,0.22635287708706328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,64,0.23872176806131998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,32,0.2404444482591417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,2560,0.31330132484436035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,65536,256,0.11550933784908718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,16384,4.0044534471299915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,65536,51200,25.238201565212673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,51200,5.409376780192058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,65536,7.213692559136285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,16384,1.3513928519354925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,12288,0.9592950608995227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,16384,4.560640123155382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,12288,3.7592442830403647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,10240,0.8617279794481066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,12288,3.5610686408148875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,10240,2.6385298834906683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,10240,2.892360899183485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,51200,12.987819247775606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,8192,0.6783502366807727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,8192,2.143085267808702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,8192,2.5603368547227645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,7168,1.8048008812798395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,7168,1.80823728773329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,6144,0.5122017860412598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,7168,0.609481758541531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,65536,16.129779391818577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,6144,1.6644550959269206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,5120,1.3740915722317164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,6144,1.589029312133789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,5120,0.4287982251909044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,4096,1.0328533384535048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,4096,0.36537599563598633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,5120,1.2745190726386177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,3584,0.3216204378339979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,3072,0.7154764599270291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,3584,0.9644578297932943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,3584,0.9018026987711588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,3072,0.28147821956210667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,3072,0.7900231149461535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,2560,0.6699288686116537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,2560,0.25105423397488064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,2048,0.5562399758232964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,2560,0.6405199898613824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,1536,0.38100266456604004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,4096,1.0217377344767253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,2048,0.20303732819027373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,1536,0.16737243864271376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,2048,0.5291857719421387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,1024,0.30094843440585667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,1024,0.13016533851623535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,768,0.2205911080042521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,1024,0.31105510393778485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,512,0.1630213393105401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,1536,0.614542219373915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,768,0.14263644483354357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,768,0.24738934304979113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,256,0.1070551143752204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,512,0.10443022516038682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,512,0.22016533215840658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,128,0.08811022175682916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,256,0.10675733619266087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,51200,19.99289279513889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,64,0.0837751097149319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,256,0.19473600387573242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,51200,32,0.10113244586520725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,128,0.19882133271959093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,64,0.1834319962395562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,32,0.21696444352467856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,51200,128,0.09115644296010335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,51200,1.8799254099527996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,65536,2.2068862915039062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,16384,0.6287964185078939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,16384,1.4166622161865234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,51200,4.006633758544922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,12288,0.3186124430762397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,16384,1.6881057951185439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,12288,1.0341164271036785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,12288,1.041470209757487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,10240,0.27548710505167645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,10240,0.9746213489108615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,10240,0.8680764304267036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,8192,0.6797742313808866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,8192,0.2526631090376112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,8192,0.7267351150512695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,7168,0.580616898006863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,65536,5.232979668511285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,7168,0.22613244586520723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,7168,0.6801253424750434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,6144,0.5236799981858995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,6144,0.17468267016940645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,5120,0.45836178461710614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,5120,0.16548977957831487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,6144,0.5445680088467068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,4096,0.3242222203148736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,5120,0.4180533356136746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,4096,0.14483643902672663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,3584,0.26423822508917916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,4096,0.3914693461524115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,3584,0.11046310928132798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,51200,7.757026672363281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,3072,0.2598373360104031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,3584,0.30828621652391225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,3072,0.09531555573145549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,2560,0.21816355652279326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,2560,0.08720266819000244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,3072,0.26678045590718585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,2048,0.16295022434658474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,65536,8.147230360243055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,2048,0.07214666737450494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,1536,0.11698844697740342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,1024,0.09777333339055379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,2560,0.2655831178029378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,1536,0.0587680008676317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,768,0.07150844732920329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,2048,0.1803377734290229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,51200,65536,29.440382215711807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,512,0.0530373321639167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,1536,0.14776178201039633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,1024,0.05077155431111654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,256,0.04177511069509718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,768,0.0447617769241333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,512,0.042430222034454346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,768,0.08629866441090901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,128,0.03653511073854234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,256,0.037478221787346735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,64,0.026912000444200303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,512,0.07240977552202013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,16384,32,0.028241776757770117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,16384,128,0.03819288810094198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,256,0.06529066960016887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,128,0.06152711311976115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,64,0.06334311432308622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,32,0.06451644500096639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,16384,1024,0.1073111097017924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,51200,1.2175529268052843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,65536,1.729502254062229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,16384,1.0457483927408855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,16384,0.3977973461151123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,12288,0.8046186765034994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,12288,0.25755378935072154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,12288,0.7909484439425998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,16384,1.2484888500637479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,10240,0.21631644831763375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,10240,0.6618462138705784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,10240,0.6897848976982964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,8192,0.5555466545952691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,8192,0.18250932958390978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,65536,4.317363739013672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,8192,0.5269680023193359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,7168,0.1526133351855808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,51200,3.041656917995877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,7168,0.5236693488226997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,6144,0.13316799534691706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,5120,0.33316177792019314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,7168,0.5786764356825087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,5120,0.11470666196611191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,6144,0.38387287987603086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,51200,4.763232760959201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,4096,0.25745511054992676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,4096,0.09514844417572021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,5120,0.40471821361117893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,3584,0.20668000645107695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,4096,0.2642737759484185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,3072,0.21260355578528511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,3584,0.24331111378139922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,3072,0.07561866442362468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,2560,0.15700356165568033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,3072,0.203712887234158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,2560,0.06754399670494927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,2560,0.1661520004272461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,1536,0.10416177908579509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,2048,0.05671199825074938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,2048,0.14539644453260633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,1024,0.07054044140709771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,6144,0.4175857702891032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,1536,0.04699466625849406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,768,0.06261689133114286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,1536,0.11221955882178412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,512,0.04155822263823615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,1024,0.0846231116188897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,768,0.03635466761059231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,256,0.028770665327707928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,768,0.06605688730875652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,512,0.0336426662074195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,512,0.0565813316239251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,256,0.030412445465723675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,256,0.05032799972428215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,65536,7.135188632541233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,128,0.029847999413808186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,2048,0.15596089098188612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,32,0.021551999780866835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,128,0.04793511165512932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,32,0.0507297780778673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,12288,64,0.049639112419552274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,1024,0.03919555412398444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,128,0.024033778243594702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,12288,3584,0.09339466359880234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,12288,64,0.020449777444203694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,16384,0.7579662005106608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,65536,1.348216904534234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,51200,0.9916889402601453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,16384,0.2802666558159722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,16384,0.9384951061672635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,12288,0.21292622884114584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,12288,0.8083964453803169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,10240,0.5941137737698025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,51200,2.6322559780544705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,10240,0.1816026634640164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,65536,3.3609608544243703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,12288,0.941832012600369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,10240,0.5994577937655979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,8192,0.5213004218207465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,8192,0.14555111196306017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,7168,0.12987911701202393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,7168,0.41343021392822266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,6144,0.4050062232547336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,7168,0.3818639914194743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,6144,0.1141857835981581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,6144,0.3288159900241428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,5120,0.3238568835788303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,5120,0.09602133433024089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,4096,0.21828888522254097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,3584,0.19238844182756212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,4096,0.10732355382707383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,5120,0.4125724368625217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,4096,0.21519200007120767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,3584,0.1945671108033922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,3584,0.07653333081139459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,3072,0.2024231089486016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,2560,0.13790844546424016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,3072,0.06426755587259929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,8192,0.4569057888454861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,3072,0.1647688945134481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,2048,0.11557777722676595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,2560,0.058993776639302574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,51200,4.841875712076823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,2560,0.1458817720413208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,1536,0.09070666631062825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,1024,0.06503378020392524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,2048,0.04850577645831638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,768,0.047813332743114896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,1536,0.049916442897584706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,2048,0.1164168914159139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,1024,0.03465244505140517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,512,0.033779554896884494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,1536,0.09558844566345215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,768,0.03887910975350274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,256,0.025251555773946974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,1024,0.07275555531183879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,128,0.018777777751286823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,512,0.029269334342744615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,65536,6.003827412923177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,768,0.05668266614278158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,256,0.027817779117160376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,64,0.02176711128817664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,512,0.048166222042507596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,10240,32,0.019279999865425956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,10240,128,0.026348445150587294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,256,0.04296977652443779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,128,0.04075555668936835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,64,0.041280001401901245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,10240,32,0.04581155710750156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,16384,0.7254906760321723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,51200,0.9699849022759331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,65536,1.2203270594278972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,16384,0.26999465624491376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,51200,2.0236665937635636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,16384,0.8255706893073188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,12288,0.19163821803198922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,12288,0.7074809074401855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,10240,0.15978311167822942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,10240,0.4191688961452908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,12288,0.8335830900404188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,8192,0.4590737554762099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,10240,0.45017867618136936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,8192,0.13076089488135442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,7168,0.2837973435719808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,65536,3.0063501993815103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,7168,0.11060888899697198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,8192,0.3639475504557292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,6144,0.26561954286363393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,6144,0.11506311098734538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,7168,0.32785601086086696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,51200,3.1493333180745444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,5120,0.24100799030727812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,5120,0.09844888581169976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,4096,0.200765331586202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,6144,0.28687911563449436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,3584,0.16983911726209852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,4096,0.07517866955863105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,5120,0.22375022040473092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,3072,0.13413066334194607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,2560,0.1170097721947564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,3584,0.06829333305358887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,4096,0.17643466260698107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,3072,0.06053600046369764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,2048,0.09453422493404812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,3584,0.1748888889948527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,3072,0.1532471047507392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,2560,0.0491831103960673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,1536,0.0690666635831197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,2048,0.04235644472969902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,1024,0.04934755629963345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,2560,0.11948000060187446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,768,0.03952444593111674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,2048,0.0964319970872667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,1536,0.03644266724586487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,512,0.029063112205929224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,1024,0.03033866816096836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,256,0.019742222295867074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,1536,0.07855200105243258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,1024,0.05890488624572754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,768,0.026259554757012263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,128,0.017052445146772597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,512,0.02334311107794444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,768,0.047084444099002414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,256,0.02236888971593645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,64,0.014097778333557976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,512,0.038895110289255776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,8192,32,0.015245333313941956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,8192,128,0.021241777473025854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,256,0.034104890293545194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,128,0.03204800022972955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,65536,4.801403469509549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,64,0.03212799959712558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,8192,32,0.03257866700490316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,16384,0.5164880222744411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,51200,0.8176328870985242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,65536,1.088362693786621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,16384,0.25120266278584796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,16384,0.6797537803649902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,51200,1.7556586795383031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,12288,0.5981929037306044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,12288,0.16659911473592123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,65536,2.290056016710069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,10240,0.39954755041334367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,10240,0.13846222559611002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,12288,0.5603546566433376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,8192,0.32324621412489146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,8192,0.11113422446780735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,10240,0.5474141968621148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,7168,0.27339288923475474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,8192,0.41419466336568195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,51200,2.8108533223470054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,7168,0.09935110807418823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,6144,0.22913156615363228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,6144,0.09395288758807713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,4096,0.14734489387936062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,7168,0.32985954814487034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,5120,0.07475733094745211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,6144,0.2401395638783773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,3584,0.13713778389824763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,5120,0.20959199799431694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,4096,0.0652444428867764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,3584,0.05607910950978597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,65536,3.4263519710964627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,3072,0.10954488648308648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,4096,0.16409511036343044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,2560,0.09531644317838882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,3584,0.13853600290086535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,3072,0.053258667389551796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,2048,0.08174933327568902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,2560,0.04452888833151924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,1536,0.05930755535761515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,3072,0.12211110856797959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,1024,0.03932355509863959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,2048,0.037131554550594754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,2560,0.1080826653374566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,768,0.033816887272728816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,1536,0.03196000059445699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,2048,0.08614577849706014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,512,0.02494222256872389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,1536,0.06895911031299166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,1024,0.02612888813018799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,256,0.01792533364560869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,768,0.02345422241422865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,1024,0.05247999893294441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,512,0.02124533388349745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,768,0.042973332934909396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,128,0.01408266690042284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,256,0.01921066641807556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,512,0.034056001239352755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,64,0.012156444291273752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,256,0.02976089053683811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,32,0.012553777959611682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,7168,5120,0.18086755275726318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,7168,128,0.019204444355434842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,128,0.02843911117977566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,64,0.028107557031843398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,7168,32,0.02843377656406826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,16384,0.4552719857957628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,16384,0.5790293481614854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,16384,0.1903537776735094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,51200,0.6937297715081109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,65536,0.8756062189737955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,12288,0.1865395572450426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,12288,0.3844906753963894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,51200,1.5037831200493708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,10240,0.4229119883643256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,12288,0.43295910623338485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,10240,0.12675110499064127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,8192,0.257388432820638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,65536,1.9519502851698134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,8192,0.09736977683173285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,7168,0.24166578716701934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,10240,0.36879200405544704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,7168,0.09375200006696914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,8192,0.384660455915663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,6144,0.21397156185574004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,6144,0.07784177859624226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,7168,0.267247994740804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,5120,0.1662613285912408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,4096,0.13294400109185114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,6144,0.2333448992835151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,51200,2.519911024305556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,4096,0.056759999858008496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,3584,0.1191662285063002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,5120,0.17755733595954049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,3584,0.05778577592637804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,3072,0.09997244675954182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,4096,0.15355555216471353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,2560,0.0857342216703627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,3072,0.0448035564687517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,3584,0.12455288569132487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,2048,0.06679377953211467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,3072,0.10640977488623725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,2560,0.09178488784366184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,2048,0.03410133388307359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,2048,0.07626133494906955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,1024,0.03586577706866794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,1536,0.028925332758161757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,1536,0.060419552856021456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,768,0.025792888469166223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,1024,0.023572444915771484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,65536,3.7992223103841147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,1024,0.04568355613284641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,512,0.027858667903476294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,768,0.02118844456142849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,5120,0.06586844391292997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,256,0.015948444604873657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,768,0.03730577892727322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,512,0.023972445064120825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,128,0.012876444392734103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,512,0.029899554120169744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,256,0.01755555636352963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,64,0.012089778151777057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,256,0.02572800053490533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,128,0.017479111750920612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,32,0.01276711126168569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,128,0.024866667058732774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,6144,2560,0.03943377733230591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,64,0.02491644521554311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,6144,1536,0.0518488883972168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,6144,32,0.02471911079353756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,16384,0.3948480023278131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,51200,0.5809209081861708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,65536,0.7060729132758247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,51200,1.1413635677761502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,16384,0.17124888632032606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,65536,1.5543431176079645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,16384,0.507879998948839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,12288,0.3689617845747206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,12288,0.13677510950300428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,10240,0.276676442888048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,10240,0.10605066352420384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,8192,0.2655751175350613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,8192,0.11181510819329156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,7168,0.21304710706075033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,10240,0.31625866889953613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,8192,0.25225599606831867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,6144,0.14730932977464464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,7168,0.10172622071372138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,7168,0.30975111325581867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,51200,2.183717303805881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,5120,0.13314578268263075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,6144,0.0670293304655287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,4096,0.10142578019036187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,6144,0.17820355627271864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,5120,0.06383466720581055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,4096,0.04874044325616625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,3584,0.08944444523917304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,65536,2.721621407402886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,3072,0.07837066385481092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,5120,0.2178808848063151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,4096,0.13658666610717773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,2560,0.0649884475602044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,3584,0.10548088947931926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,2048,0.05676622523201836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,3072,0.04170044594340854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,12288,0.35317065980699325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,3072,0.10199822319878472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,2560,0.03451022174623277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,1536,0.04273955689536201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,1024,0.028924445311228435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,2048,0.028757333755493164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,2560,0.11790755059983994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,768,0.021980444590250652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,1536,0.024302222662501868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,512,0.016741333736313712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,2048,0.0652079979578654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,1024,0.021319111188252766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,1536,0.05188888973659939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,256,0.01277511070171992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,1024,0.039911998642815485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,128,0.010168000227875179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,512,0.017078222499953378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,768,0.032482667101754084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,256,0.016538666354285348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,64,0.008950221869680617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,512,0.026260443859630164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,5120,32,0.00925066653225157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,256,0.022321777211295232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,128,0.0158915552828047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,128,0.021704889006084863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,64,0.02160444524553087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,5120,32,0.021713778376579285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,3584,0.04439555605252584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,65536,0.5724248886108398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,16384,0.36506133609347874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,51200,0.46258846918741864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,5120,768,0.019183110859658983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,51200,0.9569262398613824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,16384,0.15434755219353571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,12288,0.2580995559692383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,65536,1.279961797926161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,12288,0.10498488611645168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,10240,0.22920444276597765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,16384,0.4685804578993056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,12288,0.2836506631639269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,8192,0.18098933166927764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,10240,0.08927644623650445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,8192,0.07109066512849596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,7168,0.1585084464814928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,10240,0.2598151101006402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,8192,0.19437955485449898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,7168,0.06388622522354126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,6144,0.14207732677459717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,65536,2.0172061920166016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,5120,0.11114488707648383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,6144,0.05688799752129448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,7168,0.16935555140177408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,4096,0.08416799704233806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,5120,0.05359289050102234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,6144,0.1478497717115614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,3584,0.06784177488750882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,4096,0.039572444227006696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,5120,0.1489431063334147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,3072,0.06022400326199002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,3584,0.036455111371146307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,4096,0.10047199991014268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,2560,0.05618400043911404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,2048,0.04227377639876472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,3584,0.09370933638678657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,2560,0.028593778610229492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,3072,0.07736266983879937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,1536,0.03345688846376207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,2048,0.024915556112925213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,2560,0.06571555799908109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,1024,0.022537777821222942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,2048,0.054995556672414146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,1536,0.021360889077186584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,768,0.018431999617152743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,1536,0.04383466641108195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,1024,0.018407111366589863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,512,0.013850666582584381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,1024,0.03277955452601115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,768,0.01681866745154063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,256,0.010335110955768162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,768,0.02702933384312524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,512,0.015153777268197803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,128,0.009072889056470659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,256,0.014182221558358936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,64,0.007958222594526079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,256,0.019015999303923715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,128,0.01423111061255137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,4096,32,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,128,0.01850755512714386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,64,0.017869333426157635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,32,0.018052443861961365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,4096,3072,0.035177777210871376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,51200,1.519157303704156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,65536,0.529711988237169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,51200,0.42079999711778426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,51200,0.8053698009914823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,65536,1.0511893166436088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,4096,512,0.021977777282396953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,16384,0.3083893458048503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,16384,0.12660443782806396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,12288,0.23809867435031465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,10240,0.17682577504052055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,12288,0.09633066919114854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,16384,0.36038933859931094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,8192,0.14623467127482095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,10240,0.09290399816301133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,12288,0.2769697772132026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,10240,0.21912888685862222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,8192,0.06631555822160509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,7168,0.1270000007417467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,51200,1.2838071187337239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,8192,0.1738222175174289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,6144,0.11866488721635605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,5120,0.09100977579752605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,7168,0.061576002173953585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,6144,0.057187557220458984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,4096,0.07115911112891303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,5120,0.043785777356889516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,7168,0.19384355015224883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,3584,0.06573510832256742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,6144,0.1313733392291599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,4096,0.03718400001525879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,3072,0.06003733476003011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,5120,0.11150488588545059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,2560,0.04667377803060743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,4096,0.09665066666073269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,3584,0.03410311208830939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,2048,0.038895110289255776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,3072,0.029626667499542236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,3584,0.07977510823143853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,1536,0.029719998439153034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,2560,0.026554667287402686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,3072,0.06919999917348225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,2560,0.05899288919236925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,2048,0.022479999396536086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,768,0.016265778077973258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,2048,0.05068622363938225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,1536,0.01959733333852556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,1024,0.016200888488027785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,1536,0.03933777742915683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,512,0.013059555656380124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,1024,0.029793779055277508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,768,0.015623110863897534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,256,0.0094035557574696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,768,0.024685333172480266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,128,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,256,0.013331555657916598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,512,0.019765333996878732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,64,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,256,0.01717422240310245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,65536,1.8106142679850261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,32,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,128,0.013350221845838757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,128,0.01686311099264357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,64,0.016248888439602323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3584,32,0.016543111867374845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3584,1024,0.02036711076895396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,65536,0.4448666572570801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,65536,0.9577129152086047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3584,512,0.014107555150985718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,16384,0.2722240024142795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,51200,0.3658488856421576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,51200,0.842407120598687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,16384,0.10959733194775051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,12288,0.20516533321804473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,12288,0.10893066724141438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,10240,0.17454400327470568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,65536,1.3755359649658203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,16384,0.3163848982916938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,51200,1.107832908630371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,12288,0.22677156660291883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,10240,0.09606755442089504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,8192,0.12380266189575195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,7168,0.11693599489000107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,8192,0.05816977553897434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,10240,0.22343820995754668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,6144,0.12605422072940403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,7168,0.05199377735455831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,5120,0.0885724425315857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,6144,0.04784977767202589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,8192,0.15751288996802434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,7168,0.13594667116800943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,4096,0.06800444258583917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,5120,0.041602667835023664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,6144,0.11882578002081977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,3584,0.05942222144868639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,4096,0.03264177838961283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,3072,0.050795555114746094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,5120,0.1018986635737949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,2560,0.04068533248371548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,3584,0.02914666798379686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,4096,0.0810391108194987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,2048,0.03291199935807122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,3072,0.026501332720120747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,3584,0.0706897775332133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,1536,0.025978666212823655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,2560,0.023956444528367784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,1024,0.018745778335465323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,3072,0.06201066573460897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,2048,0.02056533263789283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,2560,0.05464355481995476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,1536,0.01835466590192583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,2048,0.04483288857671949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,512,0.011095999843544431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,1024,0.015984889533784654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,256,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,1536,0.03609866566128201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,768,0.014860444598727755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,1024,0.02706844442420536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,128,0.00700800038046307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,512,0.01346577786737018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,768,0.022291556000709534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,256,0.012729778057999082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,64,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,512,0.01825777855184343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,32,0.006566222343179915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,256,0.015573332707087198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,128,0.015184000134468079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,64,0.01481066644191742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,3072,32,0.014853333433469137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,65536,0.4133804374270969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,51200,0.319792005750868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,51200,0.6155190997653538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,65536,0.746921804216173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,3072,768,0.015127110812399121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,16384,0.2295431031121148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,12288,0.1764648887846205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,16384,0.10507822036743164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,10240,0.140301333533393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,16384,0.26052623324924046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,12288,0.19865689012739393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,51200,0.9316932890150281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,3072,128,0.01275022245115704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,8192,0.11864177385965984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,10240,0.07601422071456909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,7168,0.09652000003390843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,10240,0.17046133677164713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,8192,0.053210665782292686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,6144,0.08126488659116957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,7168,0.05010933346218533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,8192,0.1343537833955553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,5120,0.06659999820921156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,7168,0.12598933113945857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,6144,0.04041066765785217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,4096,0.05518844392564562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,6144,0.10548088947931926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,5120,0.03798933492766486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,3584,0.04708088768853081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,5120,0.08673866589864095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,4096,0.029179556502236262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,3072,0.04110044572088454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,3584,0.026039999392297532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,4096,0.06971999671724108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,2560,0.0318177772892846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,3584,0.06869600216547649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,3072,0.023427555958429973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,3072,0.05448266532686022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,2560,0.021134222547213238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,1536,0.020622221959961783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,2560,0.0466177761554718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,65536,1.1706728405422635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,2048,0.018803555104467604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,1024,0.015575110912322998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,2048,0.039000001218583845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,1536,0.016186666157510545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,768,0.01179377817445331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,1024,0.014087999860445658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,1536,0.031559109687805176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,512,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,1024,0.023732443650563557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,768,0.01344622257683012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,256,0.006911110960774952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,768,0.019824888971116807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,512,0.012431110772821637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,128,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,256,0.01182755579551061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,512,0.016210666961140104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,64,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,256,0.013873777455753751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,32,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,128,0.011696000066068439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,128,0.013587555951542325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,64,0.013136000268989138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2560,32,0.01311644415060679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2560,2048,0.027423999375767175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2560,12288,0.08520978026919895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,65536,0.32838310135735405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,16384,0.16000621848636204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,51200,0.43913687600029844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,65536,0.6815342373318143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,12288,0.11371200614505345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,51200,0.2586115466223823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,16384,0.08385244343015884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,12288,0.06453600194719103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,10240,0.09693333175447251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,8192,0.08879733085632324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,16384,0.22209866841634116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,10240,0.058435552650027804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,12288,0.19106222523583305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,65536,0.9635778003268771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,7168,0.08150311311086018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,51200,0.7489164670308431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,8192,0.04678488771120707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,6144,0.06808355781767103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,7168,0.04405422343148125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,8192,0.11398222711351182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,5120,0.05271377828386095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,4096,0.03852533300717672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,7168,0.10306133164299859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,5120,0.030361778206295435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,6144,0.08708444568845962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,3584,0.03494400117132399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,4096,0.025009777810838487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,5120,0.07540177636676364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,3072,0.03050222330623203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,4096,0.06072711282306247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,2560,0.026742221580611333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,3584,0.023006222314304773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,2048,0.02199199961291419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,3072,0.021128889587190416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,3584,0.05520177880922953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,2560,0.018970666660202872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,1536,0.018583999739752877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,3072,0.04743644595146179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,1024,0.012773333324326409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,2560,0.04099644554985894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,1536,0.015182221929232279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,2048,0.034318221939934626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,768,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,1536,0.027635556128289964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,1024,0.013224889006879596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,512,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,1024,0.020977778567208186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,768,0.01239377756913503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,256,0.005763555566469829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,768,0.01738933391041226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,512,0.011424889167149862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,128,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,512,0.01444355481200748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,256,0.01089333328935835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,64,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,256,0.012449777788586087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,6144,0.034512887398401894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,2048,32,0.0053342220683892565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,128,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,128,0.01200711147652732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,64,0.011706666813956367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,32,0.01142133358452055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,65536,0.27031556765238446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,2048,2048,0.017157332764731515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,65536,0.5254337522718642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,51200,0.36872177653842503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,16384,0.13539111614227295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,12288,0.10415466626485188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,16384,0.07043200069003634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,51200,0.21358133686913383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,65536,0.746796449025472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,2048,10240,0.14167911476559109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,16384,0.2211866643693712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,10240,0.0931511123975118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,51200,0.5881644354926215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,12288,0.061335113313463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,8192,0.06990755266613431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,10240,0.054008887873755566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,12288,0.13869955804612902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,8192,0.04169955518510607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,10240,0.11814133326212566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,6144,0.053278221024407275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,5120,0.041343109475241765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,8192,0.10301866796281602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,4096,0.03138488862249587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,5120,0.027680888772010803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,6144,0.07245866457621257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,5120,0.06064711014429728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,3584,0.031049778064092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,4096,0.02276266614596049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,4096,0.04969244533114963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,3072,0.02637422250376807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,3584,0.020616888999938965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,2560,0.022030222747060988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,3072,0.018375110295083787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,3584,0.04496888981925117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,2048,0.018923555811246235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,3072,0.03956355651219686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,2560,0.017210667332013447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,1536,0.015063999427689446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,2560,0.03408088948991563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,2048,0.015959110524919298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,1024,0.01074488874938753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,2048,0.029013332393434312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,1536,0.013447999954223633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,7168,0.06599644157621595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,768,0.008887999587588841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,1536,0.023519999451107446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,1024,0.011744000017642975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,512,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,7168,0.03695377707481384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,768,0.01129155523247189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,256,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,1024,0.018222222725550335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,6144,0.03182755576239692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,128,0.004865777575307422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,768,0.014897776974572076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,512,0.010449777874681683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,64,0.004247111164861255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,256,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1536,32,0.004634666774008009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,512,0.012446222205956778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1536,128,0.010051555931568146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,256,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,128,0.010096888575288985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,64,0.009930666950013902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,32,0.010010666317409938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,65536,0.3554640081193712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,51200,0.28169510099622935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,65536,0.21507110860612658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,16384,0.0915786690182156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,51200,0.17013688882191977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,12288,0.07428177860048082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,16384,0.05732533004548815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,10240,0.06245155466927422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,12288,0.04597066508399116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,65536,0.5877520243326823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,16384,0.16011555989583334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,12288,0.11357155111100938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,10240,0.04159733321931627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,8192,0.0503057771258884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,7168,0.04271822174390157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,6144,0.035205334424972534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,10240,0.09560266468260024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,8192,0.03049955434269375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,7168,0.027053332991070215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1536,7168,0.08904977639516194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,5120,0.03028977910677592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,8192,0.0769875579410129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,4096,0.022501332892311945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,6144,0.02334311107794444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,7168,0.06811822122997709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,5120,0.02040177749262916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,3584,0.020399999287393358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,6144,0.058563556936052114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,4096,0.017610667480362784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,5120,0.04955200023121304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,3584,0.01629599928855896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,4096,0.041414221127827965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,2560,0.015215999550289579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,3072,0.015196444259749519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,3584,0.036885334385765925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,2048,0.01314933349688848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,2560,0.01386222243309021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,3072,0.032623999648623996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,1536,0.010616000327799054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,2560,0.028055999014112685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,2048,0.012554667062229581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,1024,0.008016000191370646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,2048,0.024040000306235418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,1536,0.01142133358452055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,768,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,1536,0.019613333874278598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,51200,0.4614186816745334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,1024,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,512,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,768,0.01017155581050449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,256,0.00424799985355801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,1024,0.015192000402344598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,128,0.0038204445607132385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,512,0.00963644435008367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,768,0.012718222207493253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,64,0.003491555651028951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,256,0.009032888544930352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,512,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,32,0.003546666767862108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,256,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,1024,128,0.0090524446633127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,128,0.008630221916569604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,64,0.008278222547637092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,1024,3072,0.01791555517249637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,65536,0.2691466543409559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,1024,32,0.008348444269763099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,16384,0.07074844174914889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,65536,0.2065582275390625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,12288,0.05423733260896471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,51200,0.16510400507185194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,16384,0.059954669740464955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,16384,0.13024977842966715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,10240,0.045794665813446045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,65536,0.5038640234205458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,12288,0.045203556617101036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,8192,0.03574577636188931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,51200,0.3937431176503499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,12288,0.09860266579522027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,10240,0.0409840014245775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,7168,0.03095910946528117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,6144,0.026928888426886663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,8192,0.029064890411165025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,10240,0.08305955595440336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,7168,0.026160889201694067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,5120,0.022636445032225713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,8192,0.0671715537707011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,4096,0.018311111463440787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,7168,0.05908799833721585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,3584,0.016889777448442247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,5120,0.020232000284724765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,4096,0.017512000269360013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,5120,0.04364888866742452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,3072,0.014701333310869006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,3584,0.01627377834584978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,4096,0.03642311029964023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,2560,0.012816888590653738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,3584,0.03234044379658169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,51200,0.22772444619072807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,3072,0.015132443772421943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,2048,0.01088711122671763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,1536,0.009102221992280748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,2560,0.013916444447305469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,3072,0.02904711167017619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,2560,0.025063110722435847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,1024,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,2048,0.012739555703269111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,1536,0.011394666300879585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,768,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,2048,0.02109866672092014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,1536,0.01756888959142897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,512,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,1024,0.010626667075686984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,256,0.004192000048028098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,768,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,1024,0.01350488927629259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,512,0.009663110805882348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,768,0.011746666497654386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,128,0.0035342222286595237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,64,0.003495111233658261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,256,0.008990222381220924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,768,32,0.003536888708670934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,512,0.009751111268997192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,256,0.008216888540320927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,128,0.009004444711738164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,768,6144,0.022914666268560622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,128,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,6144,0.05230399966239929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,65536,0.19646933343675402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,64,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,768,32,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,16384,0.050013333559036255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,51200,0.15993777910868326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,65536,0.16133866045210096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,12288,0.03836177786191305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,51200,0.12914489375220403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,16384,0.045347554816140064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,10240,0.03118844495879279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,12288,0.03899200095070733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,65536,0.4200124475691054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,16384,0.10977155632442898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,8192,0.02362755603260464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,51200,0.33127466837565106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,7168,0.021242666575643752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,10240,0.03236888845761617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,12288,0.08374489016003078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,6144,0.017292444904645283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,8192,0.02278933260175917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,10240,0.0703679985470242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,7168,0.02149066660139296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,5120,0.014352000421947904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,4096,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,8192,0.05654222435421414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,6144,0.01793511046303643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,7168,0.050415999359554715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,3584,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,5120,0.016982222596804302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,6144,0.04383200075891283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,3072,0.009753777749008602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,4096,0.014077333940400017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,5120,0.03741511040263706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,2560,0.008603555460770925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,3584,0.01312888910373052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,4096,0.030795554320017498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,3072,0.012711111042234631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,2048,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,3584,0.0277128881878323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,1536,0.006492444624503453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,3072,0.02478755513827006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,2048,0.011095110740926532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,1024,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,1536,0.01037333326207267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,2560,0.021556443638271753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,768,0.0042160000238153665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,2048,0.018242667118708294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,512,0.003625777860482534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,1024,0.009534222384293875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,768,0.00906311141120063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,1024,0.011947555674446953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,256,0.0032017777363459268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,512,0.008618666893906063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,768,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,128,0.002834666727317704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,512,0.008912000391218398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,256,0.008357333640257517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,64,0.0027715555495686005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,512,32,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,128,0.008294222255547842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,256,0.007325333025720384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,128,0.006807999892367258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,65536,0.12655644946628147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,64,0.006653333289755716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,32,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,51200,0.09867822461658055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,16384,0.037632889217800565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,65536,0.16251644823286268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,12288,0.028744889630211726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,51200,0.12804444630940756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,65536,0.3409670988718669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,16384,0.04489244355095757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,10240,0.020616888999938965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,51200,0.2682150999704997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,16384,0.08996444278293186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,8192,0.014767110347747803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,12288,0.03678844372431437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,7168,0.01297155519326528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,10240,0.031888888941870794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,12288,0.0684337748421563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,8192,0.022853333089086745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,10240,0.05726755327648587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,6144,0.011533333195580376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,5120,0.010038221875826517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,7168,0.02000711030430264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,8192,0.046148445871141225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,4096,0.008423111504978603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,6144,0.017856889300876193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,7168,0.04121422105365329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,3584,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,5120,0.01608000033431583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,6144,0.036178665028678045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,3072,0.007079111205206976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,4096,0.014122666584120857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,5120,0.03121333320935567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,2560,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,3584,0.013583111266295115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,4096,0.025244444608688354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,3072,0.01239377756913503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,3584,0.023044443792766992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,2048,0.005423111220200856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,3072,0.020358221398459542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,2560,0.011751111182901593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,1536,0.004587555511130227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,2560,0.017900443739361234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,2048,0.011025778121418424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,1024,0.0038657776183552216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,512,2560,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,2048,0.015454222758611044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,1536,0.010335110955768162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,768,0.003540444291300244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,1024,0.009480888644854227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,512,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,1536,0.01293777757220798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,256,0.0028320000403457214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,768,0.009320889082219865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,1024,0.01035377797153261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,512,1536,0.015142222245534262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,128,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,512,0.00867911097076204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,768,0.009005332986513773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,64,0.002696000039577484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,256,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,256,32,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,512,0.007786666353543599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,256,128,0.008277333445019191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,256,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,65536,0.0988008909755283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,128,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,51200,0.07892266909281413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,64,0.005640888793600931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,256,32,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,16384,0.030080888006422255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,12288,0.021934222843911912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,65536,0.159734222624037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,51200,0.13195021947224936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,16384,0.046691556771596275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,10240,0.01516533394654592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,12288,0.037626667155159846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,16384,0.07928711175918579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,51200,0.23529511027865938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,8192,0.01295466638273663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,12288,0.059860441419813365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,10240,0.03334044416745504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,8192,0.023601777023739282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,10240,0.05055911011166043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,6144,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,7168,0.020768000019921195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,8192,0.041144887606302895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,5120,0.009307555854320526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,7168,0.03634488913748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,6144,0.01869333287080129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,4096,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,5120,0.016959110895792644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,6144,0.03194844391610887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,3584,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,5120,0.02718577782313029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,4096,0.01484444406297472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,3072,0.0063697778516345555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,3584,0.014088888963063559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,2560,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,3584,0.020639111598332722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,3072,0.012998222476906247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,2048,0.005044444567627377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,2560,0.012025777664449481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,3072,0.018457776970333524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,1536,0.004519999855094486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,2048,0.011314666933483548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,2560,0.01641155613793267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,65536,0.30084888140360516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,1024,0.003539555602603488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,768,0.0032142222755485107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,1536,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,2048,0.01418133411142561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,512,0.0028435554769304064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,1024,0.009401777552233802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,1536,0.01202933324707879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,7168,0.011767110890812345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,256,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,768,0.009311111436949836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,1024,0.009724444813198512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,128,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,512,0.008637333081828224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,256,0.008275555239783393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,64,0.0022177776942650476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,512,0.007290666302045186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,128,32,0.0023022221608294383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4097,128,128,0.008270222279760573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,256,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,128,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,65536,0.09557866387897068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,51200,0.07631199889712863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,16384,0.02646044393380483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,64,0.0052915554907586836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,12288,0.017638223038779367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,32,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,10240,0.013728888498412238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,4096,0.02271822260485755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,8192,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,65536,0.2998853259616428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,16384,0.07882489098442926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,51200,0.23508710331386992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,12288,0.059789333078596324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,7168,0.010844444235165914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,10240,0.050331556134753756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,6144,0.009654222263230218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,5120,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,4096,0.00702044450574451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,3584,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,8192,0.04105600052409702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,3072,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,7168,0.03616088959905837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,2560,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,5120,0.027052443888452318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,4096,0.022636445032225713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,2048,0.004541333350870344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,1536,0.0038408889538711975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,128,768,0.008341333104504479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,3072,0.018351111147138808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,1024,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,768,0.002971555623743269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,2560,0.016216889023780823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,2048,0.014119999276267158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,512,0.002549333290921317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,256,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,1536,0.011824888487656912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,128,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,1024,0.009681777821646797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,768,0.008278222547637092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,512,0.007241778075695038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,32,0.0023440000497632557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,256,0.005667555663320754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,128,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,65536,0.08934577968385483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,64,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,51200,0.07025600141949125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,32,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,16384,0.025954667064878676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,12288,0.017059556312031217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,6144,0.03147111005253262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,10240,0.011680000358157687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,65536,0.3006239997016059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,16384,0.07834666967391968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,51200,0.23467644055684408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,8192,0.010079999764760336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,64,3584,0.020541333489947848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,7168,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,12288,0.059768888685438365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,6144,0.008575999902354347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,5120,0.011030221978823343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,4096,0.009300444689061906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,10240,0.05049955513742235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,3584,0.00869333330127928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,6144,0.031601776679356895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,3072,0.007603555917739868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,5120,0.027037334111001756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,2560,0.006743999818960826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,64,64,0.0021671110557185277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,2048,0.004566222015354368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,4096,0.022657778528001573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,1536,0.0038337777886125776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,1024,0.0032115555885765287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,3584,0.020847999387317236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,768,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,3072,0.018592000007629395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,2560,0.016559110747443307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,512,0.0025333333760499954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,2048,0.01425866617096795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,1536,0.01203555530971951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,128,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,1024,0.009661333428488838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,64,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4097,32,32,0.002016000035736296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,768,0.008571555217107138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,512,0.007541333635648091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,256,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,128,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,64,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,32,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,7168,0.035982221364974976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4097,32,8192,0.04081511166360643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,16384,1.7041759490966797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,16384,4.899947696261935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,16384,6.5455110337999125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,16384,5.233527289496528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,51200,7.005918290879991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,12288,4.041726430257161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,12288,1.2038248909844291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,12288,4.044867621527778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,10240,1.0168444315592449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,12288,3.7771119011773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,10240,3.390943103366428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,10240,3.946305592854818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,8192,0.8146044413248698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,10240,3.0884727901882596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,8192,2.7361706627739797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,8192,2.6137750413682723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,8192,2.5255209604899087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,51200,15.530417548285591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,7168,2.4904621971978083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,7168,2.2058319515652127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,7168,0.7120498021443685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,6144,0.638937791188558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,7168,2.199337853325738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,6144,2.404197269015842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,5120,1.632296880086263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,6144,2.0153564876980252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,6144,1.889217800564236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,5120,0.5266373422410753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,5120,1.6865822474161785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,4096,1.2301786210801866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,4096,0.43844087918599445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,4096,1.3044808705647786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,5120,1.606827523973253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,3584,1.08896976047092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,4096,1.2485538058810763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,3584,0.40207645628187394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,3584,1.153058687845866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,3072,0.34934221373664004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,3072,1.0250310897827148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,3584,1.100495974222819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,3072,0.9793092939588758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,2560,0.746980455186632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,3072,0.9240141974555122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,2560,0.2952631049686008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,2048,0.5911422305636936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,2560,0.8414275381300184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,2048,0.25291911760965985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,2560,0.8630400233798556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,1536,0.46302043067084414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,2048,0.691431999206543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,2048,0.6351937717861599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,1536,0.20457777712080213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,1536,0.5399040116204156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,1024,0.3416417704688178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,1536,0.4869350857204861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,1024,0.16539733939700657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,768,0.23602133327060273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,1024,0.39227644602457684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,512,0.21753156185150146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,768,0.14393866062164307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,1024,0.3895413345760769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,51200,18.04509819878472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,256,0.1304088830947876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,768,0.42370133929782444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,768,0.3035502168867323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,512,0.12789244121975368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,128,0.11127732859717475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,512,0.261677344640096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,256,0.1116373274061415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,128,0.10888177818722194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,256,0.19231555196974012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,512,0.30901866488986546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,128,0.22319910261366102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,256,0.3377946747673883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,32,0.09837244616614448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,128,0.1439973380830553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,51200,24.783435397677952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,64,0.22979821099175346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,32,0.2372871239980062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,65536,64,0.1024071110619439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,51200,5.119322458902995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,65536,6.701032850477431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,16384,4.54817157321506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,51200,12.177097744411894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,16384,5.058590359157987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,16384,1.280801773071289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,65536,16.079740736219616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,12288,3.489657719930013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,16384,4.158323499891493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,12288,0.9588737487792969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,12288,3.2974185943603516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,51200,14.016241285536024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,10240,0.8030435774061414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,51200,19.958179050021702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,12288,2.9921972486707897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,10240,2.8086969587537975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,10240,2.836004469129774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,8192,0.6968071195814344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,10240,2.458173327975803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,8192,2.120374255710178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,8192,2.145575205485026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,7168,0.6016409132215712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,8192,1.9961431291368272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,7168,2.1130843692355685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,7168,1.793166266547309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,6144,1.521783086988661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,7168,1.5190596050686305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,6144,0.5930364396837022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,65536,19.247745090060764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,6144,1.6020675235324437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,5120,0.4058764510684543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,5120,1.4115431043836806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,6144,1.4609066645304363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,4096,1.058150185479058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,4096,0.4008551173739963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,5120,1.2532320022583008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,5120,1.1043857998318143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,4096,0.9986506568060981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,3584,0.8603599866231283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,4096,0.9559866587320963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,3584,0.312138663397895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,3584,0.9117084079318576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,3072,0.28622576925489635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,3584,0.8101964526706271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,65536,27.62592061360677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,3072,0.8669510947333442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,3072,0.7156249152289497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,2560,0.24033954408433703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,2560,0.6664942105611166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,2560,0.5940737724304199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,2048,0.20215111308627656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,3072,0.7295129034254285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,2048,0.5227733188205296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,1536,0.4193262259165446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,2048,0.49419201744927305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,1536,0.1667919953664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,1024,0.25508443514506024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,1536,0.4364159901936849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,1024,0.13203467263115778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,1536,0.3812471230824788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,768,0.19882133271959093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,1024,0.3709244463178847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,512,0.15509511364830866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,1024,0.2880053255293104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,768,0.11390488677554661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,768,0.24566221237182617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,256,0.10079822275373672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,2048,0.4687031110127767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,512,0.10300533639060126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,768,0.23359200689527723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,512,0.21237333615620932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,256,0.08864355749554104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,512,0.21429599655999076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,256,0.1864017777972751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,64,0.07944355408350627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,256,0.1483128865559896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,51200,128,0.08770044644673665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,32,0.07698933283487956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,128,0.17614132828182646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,51200,128,0.11304799715677898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,64,0.180839114718967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,2560,0.5898568895128038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,51200,32,0.18461422125498453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,51200,128,0.08618933624691433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,65536,2.0562720828586154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,51200,1.5172640482584636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,51200,3.820740593804253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,16384,1.5231093300713434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,16384,1.5226667192247179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,65536,4.899610731336805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,16384,0.4486995802985297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,12288,1.115049786037869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,16384,1.4043138292100694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,51200,4.495342254638672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,12288,0.3851617707146539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,12288,1.0654239654541016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,51200,6.19998762342665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,12288,0.8918364842732748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,10240,0.26565689510769314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,10240,0.8716693454318576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,8192,0.6843493249681262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,8192,0.25180800755818683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,8192,0.7022844420539008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,10240,0.7555715772840711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,10240,1.1413111156887479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,65536,5.895002577039931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,8192,0.6064462131924099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,7168,0.7228026919894748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,7168,0.19416444831424287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,6144,0.484557310740153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,6144,0.18898133436838785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,7168,0.6102586852179633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,7168,0.5141235457526313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,5120,0.3855920102861192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,6144,0.5490195486280653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,5120,0.14527289072672525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,4096,0.32266754574245876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,6144,0.44159910413953996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,4096,0.12096444765726726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,65536,9.227350023057726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,4096,0.36980888578626847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,5120,0.3769999874962701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,3584,0.25410487916734487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,4096,0.31136711438496906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,3584,0.11204532782236735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,3072,0.22695822185940215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,3072,0.09576800134446885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,3584,0.29545511139763725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,2560,0.19869511657290986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,3584,0.2713449001312256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,3072,0.24990399678548178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,2560,0.08216444651285808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,2048,0.15843466917673746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,3072,0.2808835506439209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,2560,0.2207324504852295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,1536,0.12705600261688232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,2048,0.06977955500284831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,2048,0.17542133066389295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,2560,0.20425332917107475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,1024,0.08339822292327881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,2048,0.1673626634809706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,1536,0.05707999732759264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,5120,0.4613724284701877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,1536,0.14230933454301622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,768,0.06907288895712958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,1024,0.04770311050944858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,1024,0.10578755537668864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,1536,0.1484266652001275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,768,0.04379111197259691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,1024,0.10290133290820652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,768,0.0844871136877272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,256,0.03604977660708957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,768,0.08776355452007717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,512,0.04001511136690775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,128,0.02984444300333659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,512,0.07141244411468506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,512,0.07327377796173096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,256,0.03551910983191596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,256,0.06385778056250678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,64,0.024999111890792847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,256,0.055492444170845874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,32,0.02534488836924235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,128,0.0358862214618259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,128,0.060646222697363965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,128,0.04286311070124308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,64,0.06203111012776693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,32,0.06320266591178046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,16384,512,0.05115822288725111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,65536,1.6657742394341364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,51200,1.1704577340020073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,51200,3.169844521416558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,16384,1.3072213066948783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,16384,1.3665715323554144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,16384,0.3361990981631809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,65536,4.462969038221571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,12288,0.7752541965908475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,16384,0.9957538180881076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,51200,3.4402052561442056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,51200,4.809099409315321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,12288,0.8114426930745443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,12288,0.23825510342915854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,12288,0.6808693144056531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,10240,0.21079733636644152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,10240,0.6963849067687988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,65536,4.461137983534071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,8192,0.49928797615898984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,10240,0.6890604231092664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,10240,0.5887671046786839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,8192,0.5279928843180338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,7168,0.47198846605088973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,7168,0.15203111701541477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,8192,0.4579244189792209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,7168,0.5276693238152398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,6144,0.35893956820170086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,6144,0.1303688950008816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,7168,0.39632267422146267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,5120,0.3202728960249159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,6144,0.44939645131429035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,6144,0.3445146613650852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,5120,0.11158311367034912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,4096,0.21727821562025282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,5120,0.3565351168314616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,8192,0.20584355460272896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,4096,0.09196977482901679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,5120,0.29372445742289227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,3584,0.20820177925957573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,4096,0.26013954480489093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,3072,0.1916631062825521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,4096,0.24316355917188856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,3584,0.2251919905344645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,3072,0.07462666432062785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,2560,0.15110933780670166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,3584,0.2098106675677829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,3072,0.2011528942320082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,2048,0.12233955330318874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,3072,0.18412533071306017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,2560,0.164919998910692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,1536,0.0899839997291565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,2048,0.05446844299634298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,2560,0.16032532850901285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,65536,7.3111572265625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,2048,0.14227733347151014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,1024,0.06482666730880737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,2048,0.13417955239613852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,1536,0.04589866598447164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,768,0.052561776505576245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,1536,0.10769422186745538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,1024,0.03895466526349386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,1536,0.10821777582168579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,1024,0.08180977900822957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,512,0.038498666551378034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,3584,0.0828897754351298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,1024,0.08217066526412964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,768,0.03592888845337762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,256,0.02815822097990248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,768,0.06562666760550605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,512,0.032813333802753024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,768,0.07004711363050672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,512,0.05560533205668131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,128,0.020773332979944017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,512,0.05742044581307305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,256,0.029260443316565618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,256,0.04956533180342781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,64,0.018535999788178336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,2560,0.06615200307634142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,256,0.044695999887254506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,128,0.02898755669593811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,128,0.04716444346639845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,128,0.034543110264672175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,64,0.0480968885951572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,32,0.04878488845295376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,12288,32,0.018959111637539335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,65536,1.2730817794799805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,51200,2.613405227661133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,16384,0.9144836001925998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,51200,1.0589688618977864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,65536,3.3611742655436196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,16384,0.33303109804789227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,12288,0.743657800886366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,16384,0.8931457731458875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,51200,3.9908392164442272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,12288,0.22037600146399605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,12288,0.6822035577562121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,65536,3.6973787943522134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,10240,0.5321777661641439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,51200,2.8927004072401257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,65536,5.299458821614583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,10240,0.18793600135379365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,12288,0.6700035730997721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,10240,0.5635893079969618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,16384,0.968688858879937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,10240,0.4776933458116319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,8192,0.13947021961212158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,7168,0.37323911984761554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,7168,0.12364444467756484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,8192,0.4788391325208876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,6144,0.30929777357313365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,7168,0.3719884554545085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,8192,0.39683998955620664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,6144,0.11100888914532131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,7168,0.3414417902628581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,5120,0.2265537844763862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,5120,0.0954586664835612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,6144,0.4620070987277561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,6144,0.29559556643168133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,5120,0.26579199896918404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,4096,0.07989866866005792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,5120,0.2514337698618571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,4096,0.23750933011372885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,4096,0.20516267087724474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,3584,0.06969777743021648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,3584,0.1858595609664917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,3584,0.1831351121266683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,3072,0.06212888823615181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,3072,0.1621777746412489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,2560,0.1338133282131619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,3072,0.15940354930029974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,2048,0.10807644658618504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,2560,0.056292441156175405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,2560,0.13942578103807238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,4096,0.20749955707126191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,1536,0.08566577566994561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,2560,0.13600444793701172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,2048,0.047375109460618764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,3584,0.17742578188578287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,2048,0.11539288361867268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,8192,0.4887644449869792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,2048,0.1144568920135498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,1536,0.040031999349594116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,768,0.041895111401875816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,3072,0.14611377980973986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,1024,0.033957332372665405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,1536,0.09267733494440715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,1024,0.070104890399509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,512,0.030671109755833943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,1024,0.07206577724880643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,768,0.03133066826396518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,768,0.056351111994849316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,256,0.02209866709179348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,512,0.028925332758161757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,512,0.04778133498297798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,768,0.06080889039569431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,512,0.050092445479498975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,256,0.042522665527131825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,256,0.02603200078010559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,64,0.016682666209008958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,256,0.03968000080850389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,32,0.017816000514560275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,128,0.04026844435267978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,128,0.02572533322705163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,128,0.030192000998391047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,64,0.04109244545300802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,1024,0.05714311202367147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,32,0.041834665669335254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,1536,0.09125689003202651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,10240,128,0.019926221834288705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,65536,1.107269287109375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,65536,2.619353824191623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,51200,2.524308522542318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,51200,0.9220968882242838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,16384,0.6891875796847873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,16384,0.28576623068915474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,16384,0.6942470868428549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,12288,0.524054209391276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,16384,0.8347644276089139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,51200,3.5886522928873696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,65536,2.995470258924696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,51200,2.3137457105848527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,65536,4.157997131347656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,12288,0.5604089101155599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,10240,0.42816800541347927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,10240,0.16229155328538683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,8192,0.3381093343098958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,8192,0.1159635517332289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,10240,0.47872357898288304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,12288,0.49194664425320095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,10240,0.49592002232869464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,8192,0.35327911376953125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,7168,0.11052711142434014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,8192,0.3846835560268826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,6144,0.2564302285512288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,7168,0.3159368832906087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,7168,0.285257789823744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,6144,0.09073333607779609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,5120,0.20189510451422799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,6144,0.261957327524821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,12288,0.17416177855597603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,4096,0.15901866224077013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,6144,0.24817421701219347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,5120,0.2216648923026191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,3584,0.12614399856991237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,4096,0.07417688767115276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,5120,0.20514844523535836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,3072,0.12396889262729222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,3584,0.06238310866885715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,4096,0.16802222198910186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,7168,0.2877635425991482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,3584,0.1519884400897556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,3072,0.05324622326427036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,3072,0.13384444183773464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,3072,0.13725154929690891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,2560,0.04685866501596239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,2560,0.11549154917399089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,2048,0.08396177821689182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,2560,0.11406844192081028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,5120,0.07816177606582642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,1536,0.06506399975882636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,2048,0.04019999835226271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,2048,0.09519466426637437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,4096,0.1750488943523831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,2048,0.0956160028775533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,1536,0.07587733533647326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,1536,0.07838666439056396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,768,0.03747911254564921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,3584,0.1534071101082696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,1024,0.028513779242833454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,1024,0.05717511309517754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,512,0.02594311038653056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,1024,0.060099555386437305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,768,0.025669333007600572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,768,0.04664444592263964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,256,0.017941332525677152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,768,0.05182488759358724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,512,0.022664889693260193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,512,0.038254221280415855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,128,0.014522666732470194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,512,0.04247022337383694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,256,0.020909332566791113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,256,0.033279998434914485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,64,0.013034666577974955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,256,0.03376888897683885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,128,0.02049511174360911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,32,0.013376000026861826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,128,0.03148977624045478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,128,0.025920889443821375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,1024,0.04597333404752943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,64,0.031812445984946355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,32,0.031806220610936485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,8192,2560,0.1068613330523173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,65536,1.0233928892347548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,1536,0.0337759984864129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,16384,0.6800142394171821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,65536,2.5212239159478083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,51200,0.7603395779927572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,16384,0.2517484558953179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,51200,2.955229229397244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,16384,0.7474986712137858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,51200,2.0917235480414496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,16384,0.6050755712721083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,65536,3.547992918226454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,12288,0.5254480044047037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,12288,0.16717244519127739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,10240,0.3159591092003716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,12288,0.4856302473280165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,10240,0.14359288745456272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,12288,0.45057068930731875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,8192,0.2828257878621419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,10240,0.4139697816636827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,10240,0.3957306543986003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,8192,0.11011555459764268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,8192,0.3135982354482015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,7168,0.2847484482659234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,8192,0.30075022909376353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,51200,1.9355786641438801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,7168,0.09565422270033096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,6144,0.20056354999542236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,7168,0.28963110182020396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,65536,2.7694070604112415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,5120,0.1615333292219374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,6144,0.08351822031868829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,7168,0.26212088267008465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,6144,0.235459566116333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,4096,0.11920711729261611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,5120,0.07451555464002821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,5120,0.19678578111860487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,6144,0.2238702244228787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,4096,0.05996266338560316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,3584,0.11099377605650161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,5120,0.19035111533270943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,3072,0.10679377449883355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,4096,0.15635555320315891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,3584,0.05419111251831055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,2560,0.07553599940405951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,3584,0.13713067107730442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,3072,0.04792711138725281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,3072,0.11913243929545085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,2048,0.059940444098578564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,3072,0.12251733409033881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,2560,0.04208266735076904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,2560,0.10299288564258152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,1536,0.059352888001335986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,2560,0.10546577639049953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,2048,0.037010666396882795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,2048,0.08482578065660264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,1024,0.044307556417253285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,1536,0.030255112383100722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,2048,0.08906755844751994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,1536,0.06788355774349637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,768,0.0363395545217726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,1536,0.07203377617730035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,1024,0.0255031105544832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,1024,0.05110222101211548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,512,0.026625777284304302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,4096,0.1583662165535821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,768,0.023008000519540574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,768,0.041598220666249595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,256,0.020097777247428894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,768,0.04714399907324049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,512,0.020636444290479023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,512,0.03373333480623033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,128,0.014721777704026965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,512,0.03932266765170627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,3584,0.13963288731045193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,256,0.01868355605337355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,256,0.029260443316565618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,64,0.012433778080675336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,7168,32,0.012701333396964602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,256,0.031215111414591473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,128,0.018561777141359117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,128,0.028140445550282795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,128,0.023066666391160753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,64,0.027803556786643133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,32,0.028119110398822363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,1024,0.05724000268512302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,65536,0.8071333567301432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,51200,1.5322969224717882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,51200,0.6708444489373101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,16384,0.532809787326389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,65536,2.1137777964274087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,16384,0.1949004464679294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,12288,0.3960862159729004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,16384,0.5854221979777018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,51200,1.7397306230333116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,51200,2.3713785807291665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,16384,0.7137609057956271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,12288,0.4567102326287164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,12288,0.1370097796122233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,10240,0.3551031218634711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,10240,0.11403555340237087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,65536,2.375358157687717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,8192,0.29231288697984487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,10240,0.35210310088263613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,65536,3.290294223361545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,8192,0.10056977801852757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,10240,0.33944887585110134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,7168,0.232098658879598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,8192,0.3103848828209771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,7168,0.08685511350631714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,6144,0.19500799973805746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,8192,0.25642487737867564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,7168,0.23435288005405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,7168,0.2482933203379313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,5120,0.15681421756744385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,6144,0.07440711392296685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,6144,0.22175555759006074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,4096,0.133133331934611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,5120,0.06392266352971394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,5120,0.16854222615559897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,6144,0.22497243351406523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,3584,0.1104124453332689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,4096,0.054027554061677724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,4096,0.13981066809760198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,3072,0.09938222169876099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,4096,0.15689867072635227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,12288,0.38888978958129883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,3584,0.12051822079552545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,2560,0.07175555494096544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,3584,0.12426400184631348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,3072,0.043256001340018384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,3072,0.10609600279066299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,3072,0.11275822586483425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,2048,0.06855199734369914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,2560,0.09043822023603652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,2560,0.09421422084172566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,1536,0.048792888720830284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,2048,0.0318524440129598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,2048,0.0750560031996833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,2048,0.07966222365697224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,1024,0.03439111179775662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,1536,0.02665600015057458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,5120,0.16836799515618217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,1536,0.05936800109015571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,768,0.02646044393380483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,1536,0.06516977813508776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,1024,0.045203556617101036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,3584,0.04782133301099142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,512,0.020949333906173706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,768,0.020631111330456205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,768,0.037328001525666975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,768,0.04292800029118856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,256,0.015280889140235053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,2560,0.03782844543457031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,512,0.01809777816136678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,512,0.029723554849624634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,128,0.012366222010718452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,512,0.03549155592918396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,256,0.01668444441424476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,64,0.010808000134097205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,256,0.025841777523358662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,6144,32,0.011395555403497485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,128,0.01714933415253957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,256,0.027560888065232172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,128,0.024703111913469102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,128,0.021295999487241108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,64,0.024531554844644334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,32,0.024409777588314478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,1024,0.022247999906539917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,1024,0.05048888921737671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,65536,0.704402658674452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,51200,0.5604933102925619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,16384,0.4581679768032498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,65536,1.7222550710042317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,51200,1.5234461890326605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,16384,0.15730133321550158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,16384,0.5308293236626519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,12288,0.39025065633985734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,51200,1.8659786648220484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,16384,0.5216568840874566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,12288,0.1524791055255466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,51200,1.5518444908989801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,65536,2.594867494371202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,12288,0.3578684329986572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,65536,2.0470773908827042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,10240,0.29165599081251353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,12288,0.33407288127475315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,8192,0.2346844408247206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,8192,0.09606311056349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,10240,0.3330639998118083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,7168,0.20370488696628145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,8192,0.272636440065172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,8192,0.22313690185546875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,7168,0.07480000125037299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,7168,0.2182248830795288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,7168,0.20996888478597006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,5120,0.14076088534461126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,6144,0.06399555338753594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,6144,0.17857421769036186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,6144,0.17264089319441053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,4096,0.11612533198462592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,5120,0.055233776569366455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,5120,0.15560089217291936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,4096,0.04607111215591431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,5120,0.15151377518971762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,3584,0.09616711404588486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,4096,0.11733067035675049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,10240,0.10442666874991523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,3072,0.07618133227030437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,3584,0.04111822115050422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,3584,0.10309688912497626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,10240,0.28102487987942165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,3584,0.11026044686635335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,3072,0.09029066562652588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,2560,0.07000088691711426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,6144,0.1898933384153578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,2048,0.056651552518208824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,3072,0.09603555334938897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,2560,0.032165331972969904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,2560,0.07844533522923787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,1536,0.03988177908791436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,2048,0.027266666293144226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,2048,0.06428177489174737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,2048,0.07037955522537231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,1024,0.028705779049131606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,1536,0.022777777579095628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,1536,0.051560001240836255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,1536,0.05676533116234673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,768,0.023240000009536743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,1024,0.01924977699915568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,1024,0.039114667309655085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,4096,0.1203848918279012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,512,0.016498666670587327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,1024,0.04446844591034783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,768,0.01796000036928389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,768,0.03211822112401327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,3072,0.036306666003333196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,256,0.0124195557501581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,768,0.03761955433421665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,512,0.0162435554795795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,128,0.01016088906261656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,512,0.031423999203575984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,256,0.015244444211324057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,256,0.02235466738541921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,256,0.024040000306235418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,64,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,128,0.01516177753607432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,5120,32,0.010701333483060202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,128,0.021319111188252766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,128,0.01889511115021176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,64,0.021357332666714985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,32,0.02129688858985901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,512,0.025794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,65536,0.5640293227301704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,65536,1.1762328677707248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,51200,1.2042658064100478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,16384,0.35135287708706325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,51200,0.46139732996622723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,16384,0.14280888769361708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,65536,2.1237528059217667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,65536,1.665542178683811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,16384,0.47900088628133136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,16384,0.38027111689249676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,12288,0.28100355466206867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,12288,0.10456266668107773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,51200,1.5734364191691081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,51200,1.2981173197428386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,10240,0.20445333586798775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,12288,0.294998221927219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,10240,0.09474488761689927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,8192,0.1886533366309272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,10240,0.26237156656053334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,8192,0.07232266664505005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,7168,0.16942044099171957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,10240,0.23197245597839355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,8192,0.19573599762386748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,8192,0.19032088915506998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,6144,0.13655111524793836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,7168,0.06966755787531535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,7168,0.16947555541992188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,5120,0.11261778407626682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,6144,0.05502400133344862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,6144,0.14875288804372153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,7168,0.16897688971625435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,4096,0.09264800283643936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,5120,0.04735022121005588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,5120,0.12130133310953777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,5120,0.1280826727549235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,4096,0.10259999831517537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,12288,0.2915297879113091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,3584,0.08761511246363322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,3584,0.03391644358634949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,3072,0.06980444325341119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,3584,0.09137244357003106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,3072,0.0754515528678894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,3072,0.08092621962229411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,2560,0.05474311113357544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,6144,0.14626399676005045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,2560,0.0838737752702501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,2048,0.04319733381271362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,2560,0.028267555766635474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,3584,0.0743137796719869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,4096,0.09826488627327813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,2560,0.06928088929918078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,1536,0.034480889638264976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,2048,0.023652444283167522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,2048,0.05477244324154324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,1024,0.02455111179086897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,2048,0.06059555874930488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,1536,0.020346666375796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,1536,0.048935999472936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,768,0.018928888771269057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,1024,0.01722133325205909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,1024,0.03276177909639146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,1024,0.03859644465976291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,3072,0.030788444810443457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,768,0.02734044525358412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,768,0.015871110889646742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,512,0.01408088869518704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,768,0.032292445500691734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,512,0.014726221561431885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,512,0.021939555803934734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,2560,0.06510044468773736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,256,0.010692444112565784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,512,0.026516444153255884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,128,0.010019555687904358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,256,0.013625777430004544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,256,0.018910222583346896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,256,0.02186044388347202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,1536,0.04314844475852119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,64,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,4096,32,0.00815555536084705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,128,0.013571555415789286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,128,0.017542221479945712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,64,0.01812799937195248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,32,0.01789066692193349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,4096,0.03896888759401109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,65536,0.5058853361341689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,128,0.018206222189797294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,65536,1.2778346803453233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,65536,1.7127564748128254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,51200,0.4022515614827474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,51200,1.0786409378051758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,16384,0.28234754668341744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,12288,0.20850488874647352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,65536,1.5578489303588867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,12288,0.2514471213022868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,16384,0.3561893304189046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,12288,0.10017866558498806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,51200,1.4342213736640081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,16384,0.3814995553758409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,10240,0.1747466723124186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,51200,1.2153199513753254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,8192,0.16400355762905544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,10240,0.0816826687918769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,10240,0.2131955491171943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,7168,0.14591644869910345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,8192,0.06390400065316094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,10240,0.2305377854241265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,8192,0.17067910565270317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,6144,0.12296533584594727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,7168,0.05836088789833916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,8192,0.20162399609883627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,7168,0.14955111344655356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,5120,0.09874133268992107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,6144,0.051254222790400185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,16384,0.1428364382849799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,6144,0.1295964452955458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,4096,0.07258222500483195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,6144,0.14231733481089273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,5120,0.10937422513961792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,3584,0.06974133518007067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,4096,0.08963111374113296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,4096,0.03513244456715054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,4096,0.10016266504923503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,3584,0.0318213336997562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,3072,0.05613778034845988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,3584,0.07841422160466512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,3584,0.08972622288597955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,3072,0.028314666615592107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,2560,0.047482666042115956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,3072,0.06854844755596586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,3072,0.07959377765655518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,2048,0.035790221558676824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,2560,0.025639111797014873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,7168,0.16768177350362143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,2560,0.058561775419447154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,1536,0.0278479986720615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,2560,0.06914399729834662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,5120,0.043106665213902794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,2048,0.0492213335302141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,1024,0.020074667202101815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,2048,0.05760266383488973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,1536,0.0192239996459749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,5120,0.12277421686384414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,1536,0.03917333483695984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,768,0.01719911065366533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,1536,0.047445333666271634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,12288,0.28654932975769043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,1024,0.016142222616407607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,512,0.012795555922720166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,768,0.015021334091822306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,1024,0.03673244516054789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,768,0.024470221665170457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,256,0.009657777845859528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,768,0.030233777231640283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,512,0.019780443774329293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,512,0.013906665974193148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,128,0.007720000214046902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,512,0.025277333127127752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,256,0.013106666505336761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,256,0.01716444392999013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,64,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3584,32,0.008773333496517604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,128,0.013087111214796701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,128,0.016537777251667447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,128,0.017102221647898357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,64,0.016154666741689045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,2048,0.021695110532972548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,32,0.016194666425387066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,1024,0.029361777835422095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,65536,0.44611289766099715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,65536,0.9101538128323026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,51200,0.7014017634921603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,256,0.020406223005718656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,16384,0.25943467352125377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,51200,0.3490133285522461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,65536,1.4073048697577581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,16384,0.2957528962029351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,12288,0.1669351127412584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,51200,1.0484497282240126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,65536,1.4068693584865992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,16384,0.3363946543799506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,10240,0.16297154956393772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,12288,0.08177600304285686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,12288,0.22706400023566353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,8192,0.1343866719139947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,51200,1.0881350835164387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,10240,0.07353244225184123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,12288,0.2443822224934896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,10240,0.20100622706943086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,7168,0.12158400482601589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,8192,0.15263111061520046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,10240,0.20950044525994194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,6144,0.10401422447628444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,7168,0.0520666672123803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,7168,0.13267466757032606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,8192,0.16848088635338676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,5120,0.08227199978298612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,6144,0.045421332120895386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,6144,0.11605421702067058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,16384,0.10576799843046401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,4096,0.07085244523154365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,5120,0.038760887251959905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,5120,0.09804089201821221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,3584,0.057922667927212186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,5120,0.10974578062693278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,4096,0.03198755449718899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,4096,0.0794266661008199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,4096,0.09034311109118992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,3584,0.028841776980294123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,3584,0.06977333625157674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,3584,0.0807244446542528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,8192,0.06421244144439697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,3072,0.026477333572175767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,3072,0.060581333107418485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,3072,0.07239288753933378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,2048,0.03400711218516032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,2560,0.023244443866941664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,2560,0.0518871115313636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,7168,0.15447021855248347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,2560,0.06270044379764132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,2048,0.020449777444203694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,2048,0.04351111253102621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,6144,0.1293964385986328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,2048,0.052129778597089976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,1024,0.01960444450378418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,1536,0.03522577881813049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,3072,0.049866666396458946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,1024,0.015420445137553744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,1024,0.02625333269437154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,768,0.015484443969196744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,1024,0.0341635545094808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,512,0.011371555427710215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,768,0.014460444450378418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,2560,0.039889779355790876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,768,0.022120889690187242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,256,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,768,0.02844888965288798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,512,0.017858665850427415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,512,0.013252444565296173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,128,0.008604444563388824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,512,0.023504000571038988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,256,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,256,0.01534400052494473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,1536,0.026311111119058397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,64,0.008042666647169325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,256,0.0195768889453676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,3072,32,0.007531555162535773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,128,0.012436444560686747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,128,0.015192889504962497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,128,0.016171556380059984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,64,0.014450665977266101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,32,0.0148053334818946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,1536,0.04302488764127096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,65536,0.3912488884396023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,65536,0.7511271370781792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,51200,0.6237555609809028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,16384,0.2325057718488905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,51200,0.31776711675855845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,65536,1.201441764831543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,51200,0.9161412980821398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,16384,0.27383555306328666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,1536,0.018223999275101554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,16384,0.10509333345625137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,65536,1.2161039776272244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,12288,0.1686320039961073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,51200,0.9039502673678927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,10240,0.1376497745513916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,16384,0.2940311166975233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,12288,0.0731217794948154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,12288,0.2039795584148831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,8192,0.11325955390930176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,10240,0.06795910994211833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,12288,0.2206853363249037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,7168,0.09725333584679498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,8192,0.05111911230617099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,10240,0.19338577323489717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,8192,0.13151021798451742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,6144,0.0864115556081136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,7168,0.045647998650868736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,8192,0.15106577343410915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,7168,0.11591377523210313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,5120,0.06991377804014418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,6144,0.03930933276812235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,6144,0.1001546647813585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,4096,0.05672978030310737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,6144,0.11675022708045112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,5120,0.033957332372665405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,5120,0.0849413341946072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,5120,0.09719377756118774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,4096,0.028608888387680054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,4096,0.06877777973810832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,4096,0.08163466718461779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,3072,0.03823733329772949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,3584,0.025960889127519395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,3584,0.06141510936948988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,10240,0.16804356045193145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,2560,0.03184355629814996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,3584,0.07327822181913587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,3072,0.02327644493844774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,3072,0.053595556153191455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,2560,0.020899555749363367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,3072,0.06501866711510552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,2048,0.02492888934082455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,2560,0.04646311203638712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,1536,0.021201777789327834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,2560,0.05634222428003947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,7168,0.14423733287387425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,2048,0.03905422157711453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,1536,0.01605955594115787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,2048,0.04831555485725403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,1536,0.03121333320935567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,3584,0.05186488893296984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,768,0.012057777908113269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,1536,0.03951822386847602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,1024,0.014104000396198697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,1024,0.023732443650563557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,1024,0.030048890246285334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,768,0.013395555317401886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,768,0.019816888703240287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,256,0.006912888752089606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,768,0.025826666090223525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,512,0.016329778565300834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,512,0.01238844460911221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,128,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,512,0.021736888421906367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,256,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,256,0.013815110756291283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,64,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,256,0.01830044388771057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,32,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,128,0.011726222104496427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,128,0.013235555754767524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,2048,0.018372444642914664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,128,0.013919111755159167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,64,0.013096888860066732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,1024,0.015279110934999255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,32,0.013078221844302284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2560,512,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,65536,0.31658487849765354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,51200,0.47265153461032444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,16384,0.16544711589813232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,51200,0.2548542287614611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,65536,0.9468444188435873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,51200,0.732027530670166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,16384,0.2279600037468804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,16384,0.08124978012508817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,12288,0.13045067257351345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,65536,1.0085058212280273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,51200,0.7602862252129449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,16384,0.26550666491190594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,10240,0.10963822073406643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,12288,0.06406488683488634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,8192,0.08726133240593804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,12288,0.16763910982343885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,10240,0.05671822362475925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,12288,0.19222044944763184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,10240,0.14653333028157553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,7168,0.08223288589053683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,8192,0.04393511017163595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,8192,0.11254578166537815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,10240,0.16544089052412245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,65536,0.5763057602776421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,6144,0.07121955686145358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,7168,0.04012177719010247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,5120,0.05497866537835863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,8192,0.14680533938937718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,7168,0.0992657807138231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,7168,0.11812533272637261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,6144,0.08573066525989109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,4096,0.04364711046218872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,6144,0.10235110918680827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,5120,0.028813332319259644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,5120,0.07323200172848172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,3584,0.03679644399219089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,5120,0.08870044681761001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,4096,0.025072889195548162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,3072,0.03342399994532267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,4096,0.07240355677074857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,3584,0.05195377932654487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,3584,0.06422577963935004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,2560,0.027609777119424608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,3072,0.046279999944898814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,3072,0.020658666888872784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,3072,0.057578669653998486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,2560,0.018577777677112155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,2048,0.024137778414620295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,2560,0.039573334985309176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,2560,0.050711999336878456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,6144,0.03397422366672092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,1536,0.019143111175960965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,2048,0.01664533383316464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,2048,0.03381155596839057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,1024,0.012821333275900947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,2048,0.042903112040625684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,1536,0.014894222219785055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,1536,0.02721422248416477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,1536,0.03496088915401035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,768,0.011384888655609555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,1024,0.020610666937298246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,4096,0.05822933382458157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,1024,0.027081777652104694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,512,0.008850666383902231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,3584,0.02347999976740943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,768,0.012160888976520963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,256,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,512,0.011375111010339526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,768,0.023353776997990076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,512,0.014133334159851074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,128,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,256,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,256,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,64,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,256,0.016230222251680162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,128,0.01071111112833023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,128,0.011852444046073489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,2048,32,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,128,0.012813333008024426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,64,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,32,0.011436444189813403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,1024,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,65536,0.5443449020385742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,65536,0.26786666446261936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,51200,0.35319821039835614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,768,0.017231111725171406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,16384,0.14142044385274252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,65536,0.7488906648423937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,512,0.019842666056421068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,51200,0.21187822024027506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,12288,0.10411377747853597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,16384,0.07215733660591973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,16384,0.18408532937367758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,65536,0.8431466420491537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,51200,0.5783324241638184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,10240,0.09330577982796563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,16384,0.22373599476284453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,12288,0.1408426629172431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,8192,0.07542400227652656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,10240,0.051911112334993154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,51200,0.685624016655816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,12288,0.16488977273305258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,10240,0.121960891617669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,7168,0.06163555383682251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,10240,0.13936621612972683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,8192,0.039356443617079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,8192,0.09439910782708062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,6144,0.05599911345375908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,7168,0.034874667723973594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,8192,0.11574221981896295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,5120,0.04327377676963806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,6144,0.030403554439544678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,7168,0.10209777620103623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,4096,0.03414933217896356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,6144,0.07220800055397882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,6144,0.08726311392254299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,5120,0.025853332546022203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,3584,0.03021777669588725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,5120,0.07524533404244317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,4096,0.021697777840826247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,4096,0.0500257776843177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,3072,0.026326222552193537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,3584,0.019543111324310303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,3584,0.04433244466781616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,12288,0.058733330832587353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,2560,0.022107554806603327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,3584,0.05599200063281589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,3072,0.01789422167672051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,3072,0.03941511114438375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,3072,0.049228443039788135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,2560,0.0163102216190762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,2048,0.018402665853500366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,2560,0.03436444534195794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,7168,0.08337155315611099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,2560,0.04332177837689718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,1536,0.01443288889196184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,2048,0.01572444372706943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,2048,0.028487109475665625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,1024,0.010496888723638324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,2048,0.03617066807217068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,1536,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,1536,0.023456000619464453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,5120,0.06087911128997803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,768,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,1536,0.030020445585250854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,1024,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,1024,0.017893332574102614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,512,0.007141333487298753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,4096,0.06304977999793158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,1024,0.02370133333735996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,768,0.011071999867757162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,256,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,768,0.015326221783955893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,512,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,768,0.020638222495714825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,128,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,256,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,512,0.01867377758026123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,256,0.010510221951537661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,256,0.014993778533405729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,128,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,32,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,128,0.01000533335738712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,128,0.012425777812798819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,64,0.010006222460005019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,32,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,65536,0.34598755836486816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,65536,0.21182489395141602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,51200,0.254406213760376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,65536,0.5587173567877876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,16384,0.08883644474877252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,51200,0.16834399435255262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,512,0.012452444268597497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,51200,0.4432160059611003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,65536,0.6799537870619031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,12288,0.06478933493296306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1536,64,0.004600888739029567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,16384,0.05720711416668362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,10240,0.057072891129387744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,16384,0.14355022377438015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,12288,0.04585422078768412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,51200,0.5313146909077963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,16384,0.18482577800750732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,8192,0.04534044530656603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,12288,0.11200088924831814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,7168,0.04018844498528375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,10240,0.04007822275161743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,12288,0.13851022720336914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,10240,0.09214666816923354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,8192,0.029179556502236262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,6144,0.03461511267556085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,8192,0.07428977886835735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,7168,0.02625333269437154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,8192,0.09783377912309434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,5120,0.027959111664030287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,7168,0.06554577747980754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,6144,0.0232933329211341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,7168,0.08510666423373753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,6144,0.056419557995266385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,4096,0.023744000328911677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,5120,0.020471110939979553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,6144,0.07347999678717719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,3584,0.021695110532972548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,4096,0.017627555463049147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,5120,0.0475777784983317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,3072,0.01893422173129188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,5120,0.06322222285800509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,4096,0.03933866818745931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,3584,0.016332444217469957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,3584,0.035473777188195124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,2560,0.016520000166363187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,3072,0.01500266624821557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,3584,0.04758844441837735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,2048,0.013818666338920593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,3072,0.031326221095191106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,3072,0.042063110404544406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,2560,0.013795555465751223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,1536,0.010784889260927836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,2560,0.02684177789423201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,2560,0.03571555680698819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,2048,0.012562666502263812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,2048,0.023031110564867657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,1024,0.008349333372380998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,10240,0.1172293292151557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,2048,0.030814223819308813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,768,0.007347555624114142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,1536,0.011381333072980246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,1536,0.018917333748605516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,1536,0.025843555728594463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,1024,0.010423999693658618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,512,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,1024,0.014499555031458536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,1024,0.020623111062579684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,768,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,768,0.012409778104888068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,256,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,768,0.018216888109842937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,512,0.009325332939624786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,512,0.010610666539933948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,128,0.0038622220357259116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,512,0.015475556254386902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,256,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,4096,0.053083555565940015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,256,0.008967111508051554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,64,0.003565333369705412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,1024,32,0.003564444267087512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,256,0.0137155552705129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,128,0.008979555633332994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,128,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,128,0.010962666736708747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,64,0.008363555702898238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,65536,0.2873600059085422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,51200,0.19601689444647896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,65536,0.2066568930943807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,16384,0.07307555278142293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,65536,0.504875553978814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,51200,0.16592621803283691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,51200,0.3926977846357558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,12288,0.05129955543412102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,16384,0.13162577152252197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,16384,0.05655022462209066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,65536,0.669432904985216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,10240,0.04555199874771965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,51200,0.5268124474419488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,12288,0.045102222098244556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,12288,0.09848799970414902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,8192,0.03561066587766012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,10240,0.039677331844965615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,12288,0.1365413268407186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,10240,0.08317777845594618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,7168,0.0306675566567315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,32,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,8192,0.028794666131337483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,6144,0.026289777623282537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,7168,0.025624889466497634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,8192,0.0669084456231859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,8192,0.09565866655773586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,7168,0.05905422237184313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,5120,0.022143999735514324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,6144,0.023000000251664057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,7168,0.08361955483754475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,4096,0.01868355605337355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,6144,0.051339556773503624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,5120,0.02020711037847731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,6144,0.07270577881071302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,5120,0.04364533225695292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,3584,0.016768000192112394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,5120,0.06260089079538982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,4096,0.017500445246696472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,4096,0.036398222049077354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,3072,0.015026667051845126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,4096,0.05246399839719137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,16384,0.17703378200531006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,3584,0.016206221448050607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,2560,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,3584,0.03250666790538364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,3584,0.04655999938646952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,3072,0.015136889285511441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,3072,0.028724445237053767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,2048,0.01091466678513421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,2560,0.01388177772363027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,3072,0.04142755601141188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,1536,0.00906044410334693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,10240,0.11612710687849258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,2560,0.02479644450876448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,2048,0.012471111284361945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,1024,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,2560,0.03523555397987366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,2048,0.02107111116250356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,1536,0.01144088887506061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,768,0.00599644457300504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,2048,0.02998133169280158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,1024,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,512,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,1536,0.025464889076020982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,1024,0.01346666696998808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,768,0.010052444206343757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,1024,0.020524443851576913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,768,0.011428444749779172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,512,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,768,0.017992888887723286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,128,0.0035377778112888336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,512,0.01551999979548984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,256,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,256,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,64,0.0031679999083280563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,256,0.013379555609491138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,128,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,128,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,32,0.0032106666929192017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,128,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,64,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,65536,0.2109680043326484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,32,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,51200,0.16781155268351236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,1536,0.017487999465730455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,768,256,0.0041866666740841335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,65536,0.4201200008392334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,16384,0.05438399977154202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,768,512,0.009767110976907942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,65536,0.5275537702772353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,16384,0.1103022230996026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,16384,0.04479555620087517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,51200,0.4131644566853841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,12288,0.039819555150138006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,16384,0.14277066124810114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,10240,0.03324977888001336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,12288,0.03592977921168009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,12288,0.0833137763871087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,12288,0.11034666829639012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,8192,0.024696888195143804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,10240,0.07050400310092501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,65536,0.16051377190483943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,10240,0.09414755635791355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,7168,0.021342222889264423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,8192,0.022088888618681166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,8192,0.05650933583577474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,6144,0.0188417782386144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,8192,0.07675022549099393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,7168,0.019762666689025033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,51200,0.33026311132642955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,51200,0.12810933589935303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,7168,0.05018488897217644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,5120,0.015822221835454304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,7168,0.06725866927040948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,6144,0.017550221747822232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,4096,0.013552889227867126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,6144,0.043599112166298755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,5120,0.015616888801256815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,3584,0.012347555822796293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,6144,0.05827377902136909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,5120,0.037054220835367836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,5120,0.05058755477269491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,4096,0.013855111267831592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,3072,0.010623999767833285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,4096,0.030990223089853924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,3584,0.013014222184816996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,4096,0.04266311062706841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,2560,0.009698666632175446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,3584,0.02754755483733283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,3072,0.012364444633324942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,3584,0.037441776858435735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,2048,0.0081404447555542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,3072,0.02462311089038849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,10240,0.03188622328970168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,2560,0.011533333195580376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,1536,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,3072,0.03360889024204678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,2560,0.02174844510025448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,2048,0.010876444478829702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,1024,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,2560,0.028984887732399836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,1536,0.010019555687904358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,2048,0.018366222580273945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,768,0.00462755560874939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,2048,0.025117332736651104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,1536,0.01514755520555708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,1024,0.009316444396972656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,512,0.00387999994887246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,1536,0.021447999609841242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,1024,0.012044444680213928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,768,0.008962666822804345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,256,0.0035182221068276297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,1024,0.017535999417304993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,768,0.010155555274751451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,512,0.008535111116038429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,128,0.0028782222006056043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,768,0.015487111277050443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,512,0.008935111264387766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,256,0.00814666681819492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,64,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,512,0.013787555197874704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,512,32,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,256,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,128,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,256,0.012076444096035428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,128,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,65536,0.12219822406768799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,128,0.008997333546479544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,64,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,512,32,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,16384,0.038633777035607234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,65536,0.15829155180189344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,51200,0.12706043985154894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,65536,0.34476176897684735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,51200,0.26739467514885795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,12288,0.02920444475279914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,16384,0.09012889199786717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,16384,0.04454400142033895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,65536,0.5107608901129829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,16384,0.13797777228885227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,51200,0.4015973409016927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,10240,0.022379555635982092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,12288,0.03552533189455668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,12288,0.0681155522664388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,8192,0.017279111676745944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,10240,0.031567109955681696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,12288,0.10714399814605713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,10240,0.05699466996722751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,7168,0.015573332707087198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,8192,0.021591110361946955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,10240,0.09108355310228135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,8192,0.04634755518701342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,6144,0.013561777770519257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,7168,0.019520888725916546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,51200,0.09858666525946723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,7168,0.041238221857282854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,5120,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,7168,0.06484178039762709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,6144,0.03605422377586365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,6144,0.01735822194152408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,4096,0.009687110781669617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,5120,0.01551555593808492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,6144,0.056681778695848256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,3584,0.009182222187519073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,5120,0.030722667773564655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,4096,0.013845333622561561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,5120,0.04909155435032315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,4096,0.025656888882319134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,3072,0.008271999657154083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,4096,0.04031555520163642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,3584,0.01295466638273663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,3584,0.022879999544885423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,2560,0.0075137780772315125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,3584,0.035629332065582275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,3072,0.0122079998254776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,2048,0.006055999961164262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,3072,0.0203475554784139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,2560,0.011704888608720569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,3072,0.03189422355757819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,1536,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,2560,0.017886221408843994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,2048,0.010951999988820819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,1024,0.004522666749027041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,8192,0.07394399907853869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,1536,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,2048,0.024363555841975745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,1536,0.012972444295883179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,768,0.0038897775941424896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,1536,0.020402666595247056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,1024,0.01033333357837465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,512,0.003488000068399641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,1024,0.01687022215790219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,768,0.00902133352226681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,768,0.009033777647548253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,256,0.002830222249031067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,768,0.01514933341079288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,512,0.008345777789751688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,512,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,128,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,256,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,512,0.013168888787428537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,256,0.011397333608733283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,64,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,128,0.008029333419269985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,128,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,256,32,0.002507555608948072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,128,0.008967999782827165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,65536,0.09378133217493693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,2048,0.015536889433860779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,32,0.0058266665372583605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,1024,0.009332444104883406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,51200,0.07533955574035645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,65536,0.15831555260552302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,65536,0.30104266272650826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,16384,0.028471999698215063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,51200,0.12650133503807917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,65536,0.5056346787346734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,51200,0.23487377166748047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,256,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,12288,0.021898667017618816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,16384,0.04513422316975064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,51200,0.3972951041327582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,16384,0.07875111367967394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,10240,0.015822221835454304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,12288,0.03619644376966689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,16384,0.13675822152031794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,12288,0.05982310904396904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,8192,0.013110222087966071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,2560,0.028107557031843398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,12288,0.10628622108035618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,7168,0.011924444801277585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,10240,0.05054222212897407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,10240,0.03168711066246033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,8192,0.02221244408024682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,10240,0.08994933631685044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,6144,0.01053066634469562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,8192,0.04100177685419718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,7168,0.02032533288002014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,7168,0.036398222049077354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,5120,0.009006222089131674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,7168,0.06436000267664592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,6144,0.01832266648610433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,6144,0.031823111904992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,4096,0.007624000310897827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,6144,0.056127998563978404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,5120,0.0163128889269299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,5120,0.027216888136333887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,3584,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,4096,0.014366222752465142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,5120,0.04786755641301473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,4096,0.022673777408070032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,3072,0.006395555618736479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,4096,0.0392542216512892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,3584,0.013401777380042605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,3584,0.020626667473051283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,2560,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,256,64,0.0058293334311909145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,3072,0.012479999827014076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,3072,0.01843644513024224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,2048,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,2560,0.01182933317290412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,2560,0.016187555260128446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,2560,0.027789334456125896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,8192,0.07275378041797213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,2048,0.014129777749379476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,1536,0.004228444563017951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,2048,0.01106844428512785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,1024,0.003543111185232798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,2048,0.023592000206311543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,1536,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,1536,0.011933333343929715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,768,0.003175111073586676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,1536,0.020260444945759244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,1024,0.009393778112199571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,1024,0.00964355551534229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,512,0.0028000000036425064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,1024,0.01648533344268799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,768,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,768,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,256,0.002464888824356927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,512,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,768,0.014244443840450711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,512,0.007381333245171442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,3584,0.0352862212393019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,512,0.012368888490729861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,256,0.00831200016869439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,256,0.005886222339338726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,3072,0.03162311183081733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,64,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,32,0.002258666687541538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,128,0.008164444731341468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,256,0.011381333072980246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,128,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,65536,0.08636177910698785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,128,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,51200,0.0702471137046814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,16384,0.027104000250498455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,64,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,12288,0.01827822294500139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,10240,0.014984889162911309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,8192,0.012182222472296821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,16384,0.07866222328609891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,65536,0.2990666760338677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,12288,0.059750219186147056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,7168,0.010851555400424533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,10240,0.05048533280690511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,6144,0.009402666654851701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,5120,0.008329778081840938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,8192,0.04097422295146518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,4096,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,3584,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,7168,0.03598755598068237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,6144,0.03165155649185181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,3072,0.0058808889653947614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,5120,0.027104000250498455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,2560,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,128,128,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,2048,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,4096,0.022631999519136217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,1536,0.003863111138343811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,3584,0.02051999999417199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,1024,0.00318755561278926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,768,0.0029768887907266617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,3072,0.01828977796766493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,2560,0.016161777906947665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,512,0.0025439999169773525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,2048,0.014124444789356656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,256,0.002491555487116178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,1536,0.011811555259757571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,128,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,64,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,1024,0.009667555491129557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,768,0.008329778081840938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,64,32,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,128,32,0.0053671110007498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,65536,0.08681599961386786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,512,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,256,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,128,0.005404444618357553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,64,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,51200,0.06948622067769368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,16384,0.025067556235525344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,32,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,12288,0.017082666357358296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,64,51200,0.2351235548655192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,10240,0.011396444506115384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,8192,0.009947555760542551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,7168,0.009129777550697327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,16384,0.07841155264112684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,12288,0.059803558720482715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,65536,0.2998746765984429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,51200,0.23447110917833117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,6144,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,10240,0.05034133460786608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,8192,0.040756444136301674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,5120,0.010775110787815519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,4096,0.009481777747472128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,3584,0.008647110727098253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,7168,0.03604266709751553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,2560,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,6144,0.03139288889037238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,5120,0.0270453327231937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,4096,0.02251822253068288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,2048,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,1536,0.0038968887594011095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,1024,0.0032222221295038858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,3072,0.01885066595342424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,768,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,2560,0.016570667425791424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,512,0.0026337777574857077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,2048,0.014477334088749357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,1536,0.012067555553383298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,256,0.002305777743458748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,1024,0.009667555491129557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,128,0.0021280000607172647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,768,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,64,0.001871111078394784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,32,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,512,0.00742222203148736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,256,0.0058559998869895935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,128,0.005434666656785541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,64,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,32,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4096,32,3072,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4096,32,3584,0.021055999729368422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,16384,1.6386506822374132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,12288,1.1563893424140084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,12288,3.9296001858181424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,12288,3.9405004713270397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,16384,5.258842468261719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,16384,5.521251678466797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,10240,0.9912480248345269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,8192,0.8077804247538248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,51200,6.554545084635417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,10240,3.543696085611979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,10240,3.150737762451172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,8192,2.6657422383626304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,8192,2.463144090440538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,7168,0.6873777707417806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,6144,0.6058079931471083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,6144,1.8939136928982203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,7168,2.1757386525472007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,6144,1.8344239128960504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,5120,1.6879901885986328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,5120,0.5116346677144369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,5120,1.5246426264444988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,7168,2.3344809214274087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,4096,0.4377048810323079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,4096,1.3452168570624456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,4096,1.1974497901068795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,3584,0.3920453389485677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,3584,1.240923563639323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,3072,0.34577155113220215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,3584,1.0387182235717773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,3072,1.0356684790717232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,2560,0.8479502465989855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,2560,0.3019671175214979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,3072,0.9428124957614474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,2048,0.647507561577691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,2048,0.28461511929829914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,2560,0.9313404295179578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,1536,0.47361331515842015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,2048,0.6660773489210341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,1536,0.23152444097730848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,1024,0.3500720130072699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,1536,0.49812979168362087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,1024,0.17301689253913033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,768,0.24800801277160645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,1024,0.37495022349887425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,512,0.18303910891215006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,768,0.1564613315794203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,256,0.14766132831573486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,768,0.3184044361114502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,512,0.14943822224934897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,512,0.2722195519341363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,128,0.11242400275336371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,256,0.11806844340430366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,64,0.10238755411571926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,65536,128,0.13951733377244738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,256,0.251979562971327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,32,0.10369510783089532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,128,0.21810932954152426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,64,0.22215822007921007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,32,0.22468177477518717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,65536,51200,15.521331787109375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,16384,3.9102380540635853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,51200,4.86006842719184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,65536,6.5673705206976996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,16384,1.2651457256740994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,65536,51200,23.58866034613715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,12288,0.9375190734863281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,16384,4.383820427788629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,12288,3.4475911458333335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,10240,0.7742035653856066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,12288,3.2387903001573353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,51200,12.338333977593315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,10240,2.520596398247613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,10240,2.872902128431532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,8192,0.6374266412523057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,7168,0.5551733440823025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,8192,1.9929048750135634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,8192,2.3206159803602433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,7168,1.8688710530598958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,65536,15.579374525282118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,7168,1.738220426771376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,6144,0.4876142077975803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,6144,1.537136925591363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,5120,0.42278043429056805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,6144,1.5067724651760523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,5120,1.2563919491238065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,4096,1.0636204613579643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,4096,0.3410009013281928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,3584,0.3201057646009657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,3584,0.8316719796922473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,5120,1.177777820163303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,4096,0.9615155325995551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,3072,0.7932106653849283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,3072,0.3075617684258355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,3584,0.8473333252800835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,2560,0.6208782196044922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,2560,0.2595226764678955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,3072,0.7323662439982096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,2048,0.5103804270426432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,2560,0.6097057660420736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,2048,0.2263582282596164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,1536,0.3971857759687636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,2048,0.5009493298000759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,1024,0.2490302191840278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,1536,0.17882311344146729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,768,0.21575466791788736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,1536,0.3917386796739366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,768,0.13212888770633274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,512,0.16755733225080702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,1024,0.31264977984958225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,768,0.2514737712012397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,256,0.12385422653622097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,512,0.12198221683502197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,512,0.19719199339548746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,128,0.09614400068918864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,51200,18.850216335720486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,256,0.09022133217917548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,64,0.08720088667339748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,256,0.17922488848368326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,128,0.08695555395550197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,51200,32,0.08031911320156522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,128,0.17702310615115693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,32,0.17872799767388237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,64,0.17715377277798125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,51200,1024,0.13751199510362414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,16384,1.2196275922987196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,51200,1.5736488766140406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,65536,2.03420172797309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,51200,65536,22.429779052734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,16384,0.49037424723307294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,16384,1.3815946578979492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,12288,0.30216799841986763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,12288,1.1353573269314237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,51200,3.7316373189290366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,10240,0.2679182158576118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,12288,0.9877333111233182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,10240,0.9150400161743164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,8192,0.6381982167561849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,8192,0.2074026664098104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,10240,0.8429724375406901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,65536,4.868744744194879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,7168,0.18470399909549287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,7168,0.6571804682413737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,8192,0.663865778181288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,6144,0.42645955085754395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,6144,0.16050488419002956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,7168,0.577239990234375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,5120,0.13793867164187962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,5120,0.40668355094061953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,6144,0.4866204261779785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,4096,0.31902578141954213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,5120,0.39676711294386124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,4096,0.1150968869527181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,3584,0.2983662287394206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,3584,0.10592355330785115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,4096,0.31435733371310765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,3072,0.2295466793908013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,3584,0.27412711249457467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,2560,0.18987377484639487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,3072,0.09341333309809367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,2560,0.08163022332721286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,3072,0.25614576869540745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,2048,0.15629600154028997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,2560,0.22058844566345215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,1536,0.10975555578867595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,51200,6.366635640462239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,2048,0.06910755236943562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,1024,0.082205335299174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,2048,0.16694488790300158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,1536,0.056233776940239795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,768,0.05933244360817803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,1024,0.046748446093665234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,512,0.047579556703567505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,1536,0.15406044324239096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,1024,0.11386044820149739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,256,0.03554844525125291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,768,0.04404888881577385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,512,0.03980799847178989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,768,0.08007999923494127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,128,0.03175288769933913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,256,0.037830223639806114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,64,0.02680177821053399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,16384,32,0.027792887555228338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,16384,128,0.03484444485770331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,256,0.06332355737686157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,128,0.05861688984764946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,64,0.060177776548597545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,32,0.05975111325581869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,65536,8.465963575575087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,16384,512,0.06850488980611165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,65536,1.529865794711643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,16384,1.0429795583089192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,51200,1.1229449378119576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,16384,1.0300959481133354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,16384,0.40804444419013125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,12288,0.2403057681189643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,12288,0.8252204259236654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,51200,2.8949936760796438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,12288,0.9589369032118055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,10240,0.21345244513617623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,10240,0.7220408651563855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,8192,0.5111546516418457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,10240,0.6159520149230957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,65536,3.7054676479763455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,8192,0.1619688934750027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,7168,0.14533955521053737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,7168,0.5585386488172743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,8192,0.48563824759589297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,6144,0.3692382176717122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,6144,0.12776177459292942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,6144,0.36940444840325254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,7168,0.4684453540378147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,5120,0.12359644307030572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,4096,0.22667733828226724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,5120,0.3036159939236111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,4096,0.08941600057813857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,51200,4.43012449476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,3584,0.2334666781955295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,4096,0.23917778333028158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,3072,0.17414400312635633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,3584,0.08088444338904487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,2560,0.1484773291481866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,3584,0.20973332722981772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,3072,0.08499022324879964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,2048,0.1156453291575114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,3072,0.19484621948666045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,1536,0.09396888812383015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,2048,0.062414222293429904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,2560,0.16003110673692492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,1024,0.06893155309889051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,2048,0.12865777810414633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,1536,0.0470728874206543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,768,0.05679021941290962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,1536,0.10342488686243693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,1024,0.03886222177081638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,512,0.039862222141689725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,1024,0.07873244418038262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,5120,0.3498053285810683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,256,0.027944889333513048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,768,0.035008000002966985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,512,0.0325271122985416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,128,0.023401778605249193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,65536,6.063117133246528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,64,0.023544889357354905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,256,0.029080000188615587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,256,0.047820445564058095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,128,0.028741333219740126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,12288,32,0.020435555113686454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,128,0.04537511203024122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,64,0.0524160001013014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,12288,2560,0.06305511130226983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,32,0.04617866542604235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,768,0.06325599882337782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,12288,512,0.05504177676306831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,65536,1.2722337510850694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,16384,0.7183804512023926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,51200,0.9688302146063911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,16384,0.2654053370157878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,51200,2.4714505937364364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,16384,1.1086888843112521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,65536,3.0610302819146047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,12288,0.20255911350250244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,12288,0.6501919958326552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,10240,0.5247502326965332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,10240,0.17169866296980116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,8192,0.16467911667293972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,8192,0.43454665607876247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,10240,0.5255937576293945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,8192,0.4122755527496338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,7168,0.36257243156433105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,7168,0.16494399971432155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,51200,3.7333475748697915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,6144,0.30005333158704967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,6144,0.10816711187362671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,7168,0.41200889481438535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,12288,0.7694782151116265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,5120,0.3386524518330892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,5120,0.10149866342544556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,6144,0.3064168824089898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,5120,0.273784875869751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,4096,0.07524355252583821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,3584,0.17463466856214735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,3584,0.06975911060969035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,4096,0.2047333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,3072,0.14586577150556776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,3072,0.07006844547059801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,2560,0.12429955270555283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,3584,0.18644355403052437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,2560,0.05674844317966037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,2048,0.10359822379218207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,1536,0.07731466823154025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,2560,0.13308355543348524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,1024,0.051070223251978554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,2048,0.11314933829837376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,1536,0.03997333182228936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,1536,0.0907848874727885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,768,0.04253333475854662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,1024,0.03388888968361749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,512,0.03081688947147793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,1024,0.06696533494525485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,768,0.03087466624048021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,256,0.022122666239738464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,4096,0.19439289304945204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,512,0.028367112080256145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,512,0.045560889773898654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,65536,4.703366173638238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,256,0.025811556312772963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,256,0.040844443771574236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,64,0.016931555337376065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,32,0.017991110682487488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,128,0.024879111184014216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,128,0.038605332374572754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,3072,0.15430843830108643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,64,0.03900266687075297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,10240,2048,0.04675199919276767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,32,0.039388444688585066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,10240,768,0.05572266711129082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,10240,128,0.019242667489581637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,51200,0.9060559802585177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,65536,1.202460394965278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,16384,0.6419519848293728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,16384,0.6944613456726074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,16384,0.29998577965630424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,12288,0.5267519950866699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,51200,2.1109661526150174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,12288,0.174873775906033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,10240,0.5020542144775391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,65536,2.5802249908447266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,12288,0.6352302233378092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,8192,0.33143289883931476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,10240,0.14915022585127088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,10240,0.4792453447977702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,7168,0.27835022078620064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,8192,0.1158177720175849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,7168,0.10572266578674316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,6144,0.24930222829182944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,6144,0.09082488881217109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,8192,0.386973328060574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,5120,0.21147378285725912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,51200,2.9877813127305775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,7168,0.4072995450761583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,4096,0.1557191080517239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,6144,0.25159645080566406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,5120,0.08067999945746528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,4096,0.07200444406933255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,3072,0.12105600039164226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,3584,0.05955466959211561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,5120,0.22453512085808647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,4096,0.17486843797895643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,2560,0.09708088636398315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,3072,0.053991112444135875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,3584,0.1530088848537869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,2048,0.08083733585145739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,3072,0.12782843907674155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,2560,0.047206223011016846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,65536,3.880091349283854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,1536,0.05837777588102552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,1024,0.04563022322124905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,2048,0.039721777041753135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,2048,0.09505510992474026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,1536,0.03716177741686503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,768,0.03232266505559286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,1536,0.07214933633804321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,512,0.023976888921525743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,1024,0.027520888381534155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,768,0.02494844463136461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,1024,0.05494933327039083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,256,0.017792888813548617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,768,0.04455111092991299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,512,0.021869333253966436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,128,0.015936889582210116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,512,0.03631911012861464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,256,0.020602666669421725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,3584,0.13861332999335393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,64,0.013736888766288757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,256,0.03142311175664266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,8192,128,0.02033955521053738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,8192,32,0.01443288889196184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,128,0.02993688980738322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,64,0.03038311004638672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,32,0.029814223448435467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,8192,2560,0.1112168894873725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,16384,0.5475031005011665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,65536,0.9748018052842883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,51200,0.7866195572747124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,16384,0.2737769020928277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,51200,1.8635084364149306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,12288,0.6624755329555935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,12288,0.44750134150187176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,12288,0.18602222866482207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,65536,2.1541947258843317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,16384,0.6763386726379395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,10240,0.13136888874901667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,10240,0.39185510741339785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,8192,0.1055226657125685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,8192,0.3054995536804199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,7168,0.24399643474155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,7168,0.09556977616416083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,6144,0.2705422242482503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,10240,0.4722408718532986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,8192,0.3181448777516683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,51200,2.5256809658474393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,6144,0.08303288618723552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,4096,0.1378879944483439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,5120,0.23132265938652885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,7168,0.39509420924716526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,5120,0.07057422399520874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,6144,0.2314364380306668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,3584,0.1606337759229872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,4096,0.06371821959813435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,5120,0.18418399492899576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,3584,0.053857776853773326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,3072,0.11427911122639973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,2560,0.08844266997443305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,4096,0.14855378203921848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,3584,0.1353706651263767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,65536,3.404311074150933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,2048,0.07172355386945936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,2560,0.043508443567487926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,1536,0.058396445380316846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,1024,0.03984266519546509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,3072,0.1774515575832791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,1536,0.031082666582531396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,2048,0.08066399892171223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,768,0.03230488962597317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,1536,0.064919114112854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,1024,0.03760000069936117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,512,0.023060444328520033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,1024,0.048844443427191846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,768,0.023064000738991633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,256,0.016597333881590102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,512,0.020303999384244282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,768,0.03988088832961188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,128,0.013463111387358772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,512,0.033033778270085655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,256,0.018479110466109384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,64,0.01221422188811832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,256,0.02854222059249878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,128,0.018565333551830716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,128,0.02702755563788944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,64,0.026751110951105755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,32,0.026536888546413843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,3072,0.04867733187145657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,7168,2560,0.09832000070148045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,7168,2048,0.03711822297837999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,16384,0.4101431104871962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,7168,32,0.012802666260136498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,51200,0.648913754357232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,16384,0.18282932705349395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,51200,1.3506160312228734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,16384,0.5592915746900771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,65536,1.8920417361789281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,12288,0.4200924502478705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,12288,0.14371644126044378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,12288,0.3828551239437527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,10240,0.31335467762417263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,10240,0.1135093371073405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,8192,0.23289510938856336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,65536,0.8782569037543403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,10240,0.44389067755805117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,51200,2.1598826514350042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,8192,0.09118933147854275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,7168,0.22760534286499023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,6144,0.17664710680643717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,7168,0.08136000235875447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,8192,0.33962223264906144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,5120,0.13779732916090223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,6144,0.07887111107508342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,6144,0.1958764394124349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,65536,3.0754328833685984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,5120,0.06350488795174493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,4096,0.12069600158267552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,3584,0.10577866766187881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,5120,0.16399555736117893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,4096,0.05254222287072075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,3072,0.08963644504547119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,3584,0.05051911208364698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,4096,0.1455635494656033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,2560,0.08263555500242445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,3584,0.11545066038767497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,3072,0.04148888919088576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,2048,0.06062755319807264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,2560,0.03747466537687514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,3072,0.10093510813183254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,1536,0.05116622315512764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,2048,0.03126044405831231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,1024,0.03207733233769735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,2560,0.08879733085632324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,768,0.026935110489527386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,1536,0.026745777991082933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,1024,0.023676445086797077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,512,0.020336889558368258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,1536,0.058119999037848584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,1024,0.0432515541712443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,768,0.019786667492654588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,256,0.014119999276267158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,512,0.018710222509172227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,128,0.012190221912331052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,768,0.0354364448123508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,512,0.029077334536446467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,64,0.011976000335481433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,256,0.016872889465755887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,7168,0.22250665558709037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,256,0.024737776981459722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,6144,32,0.01276177747382058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,6144,128,0.016515556308958266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,128,0.02365511159102122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,64,0.023697778582572937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,32,0.023412444525294836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,16384,0.3350497881571452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,6144,2048,0.07163200113508436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,65536,0.7012355592515734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,51200,0.551879988776313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,16384,0.16007467110951742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,51200,1.151407135857476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,12288,0.3427911069658067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,12288,0.12071732680002849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,16384,0.45557064480251735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,10240,0.298363553153144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,8192,0.2105191151301066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,10240,0.10289155774646336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,12288,0.4024835427602132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,65536,1.758387565612793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,7168,0.21054489082760283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,8192,0.08057866493860881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,10240,0.40255289607577854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,7168,0.09373778104782104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,6144,0.17772177855173746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,7168,0.19320444266001383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,8192,0.2926391230689155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,5120,0.1753795544306437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,6144,0.06601422362857394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,51200,2.099517398410373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,4096,0.12619466251797146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,6144,0.16932977570427787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,5120,0.05413066678576999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,3584,0.10217422246932983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,4096,0.04630577895376417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,3072,0.07675821913613214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,65536,2.4767191145155167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,5120,0.1588675578435262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,3584,0.03984977801640829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,2560,0.06646222538418241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,4096,0.11553422609965007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,2048,0.053189333942201406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,3072,0.03611822260750665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,3584,0.10462044345008002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,1536,0.041768001185523145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,2560,0.03524533245298598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,1024,0.027709333433045283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,3072,0.10372089015112983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,2048,0.02737422287464142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,2560,0.07748888598548041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,1536,0.022786666949590046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,768,0.022320000661744013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,512,0.016521778371598985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,2048,0.06185599830415514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,1024,0.01906399925549825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,1536,0.05078577664163378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,256,0.012279999752839407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,768,0.01756444407833947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,1024,0.038133333126703896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,128,0.00959022260374493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,512,0.016177778442700703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,768,0.031075557072957356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,64,0.008564444051848518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,512,0.024890666206677754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,5120,32,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,128,0.014910222755538093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,256,0.021344888541433547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,128,0.0211191111140781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,64,0.020742222666740417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,5120,32,0.020576889316240948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,65536,0.5582035382588705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,51200,0.4451324674818251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,51200,0.9604266484578451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,65536,1.2068559858534071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,5120,256,0.01515111161602868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,16384,0.14289511574639216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,12288,0.2391288810306125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,16384,0.39267733361985946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,12288,0.10284177462259929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,10240,0.20172177420722115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,12288,0.2726835674709744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,51200,1.4928408728705511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,10240,0.08650222089555527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,8192,0.1655342181523641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,7168,0.1380808883243137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,10240,0.2300044430626763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,65536,1.956857681274414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,8192,0.07198488712310791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,6144,0.13001244597964817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,7168,0.06993155346976386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,8192,0.1814426713519626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,5120,0.10289866394466823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,6144,0.05900355842378405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,4096,0.0826915568775601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,7168,0.15917332967122397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,5120,0.04775289032194349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,6144,0.13990399572584364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,16384,0.32781955930921763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,3584,0.07289955351087782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,4096,0.04036444425582886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,3072,0.05983555316925049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,5120,0.1157191064622667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,4096,0.0952133337656657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,2560,0.052742222944895424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,3584,0.033547557062572904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,2048,0.04131644301944309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,3072,0.029824889368481104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,3584,0.08229244417614408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,2560,0.027365333504146997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,3072,0.07232266664505005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,2048,0.02342133389578925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,1024,0.02370400064521366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,2560,0.06221510966618856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,1536,0.019941333267423842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,2048,0.05300977826118469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,1024,0.01717511150572035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,512,0.013850666582584381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,1536,0.04158755474620395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,256,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,1024,0.031332443157831825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,768,0.026183999247021143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,512,0.01443911095460256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,128,0.008788444101810455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,256,0.013594666288958656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,512,0.021012443635198805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,64,0.0076248885856734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,256,0.018233777748213876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,32,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,128,0.013439111411571503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,128,0.01755199995305803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,64,0.017518222332000732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,4096,32,0.01754666699303521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,1536,0.03290133343802558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,4096,768,0.01862844493654039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,65536,0.49626933203803164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,4096,768,0.015485333071814643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,51200,0.7400737868414985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,16384,0.2886871231926812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,51200,0.3865484396616618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,16384,0.11853510803646511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,65536,1.1780897776285808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,12288,0.20659821563296846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,16384,0.3767395549350315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,12288,0.08881688780254787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,10240,0.2062657806608412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,12288,0.2375591066148546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,10240,0.08856266736984253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,8192,0.14541867044236925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,7168,0.1296773354212443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,51200,1.2269182205200195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,10240,0.29666222466362846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,65536,1.6294053395589192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,6144,0.10903111431333755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,8192,0.1966195503870646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,7168,0.055451555384529956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,7168,0.15527289443545872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,5120,0.08399555418226455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,6144,0.05328266819318136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,4096,0.07130489084455702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,5120,0.04118044508828057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,3584,0.0595911078982883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,6144,0.1863751146528456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,4096,0.03829333186149597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,5120,0.10723733239703709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,3072,0.05442577931616041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,4096,0.08897155523300171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,3584,0.03131644593344794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,2560,0.04249511162439982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,3584,0.07495288716422187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,2048,0.03486844566133287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,2560,0.02546666728125678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,3072,0.06727288828955756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,1536,0.02896089024013943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,2048,0.02241955531968011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,2560,0.058661335044436984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,1024,0.019511111908488803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,2048,0.04749600092569987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,1536,0.018579555882347953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,768,0.016199111938476562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,1024,0.01588355501492818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,1536,0.037957333856158786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,512,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,8192,0.0758533345328437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,768,0.01477866702609592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,256,0.008996444443861643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,1024,0.028414222929212783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,512,0.013794666363133324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,128,0.007746666669845581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,64,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,256,0.013082666529549493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,512,0.019088000059127808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3584,32,0.007253333098358578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,128,0.013106666505336761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,256,0.016548444827397663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,128,0.016145777371194627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,64,0.0155395550860299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,32,0.016217778126398723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3584,3072,0.027907556957668726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,65536,0.43267732196384007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,16384,0.2536337905459934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,51200,0.34566932254367405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,51200,0.6375359959072536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,65536,0.9502062267727323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,12288,0.1876035531361898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,16384,0.10401778088675605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3584,768,0.023591111103693645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,12288,0.08616799778408474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,16384,0.31214133898417157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,10240,0.18056978119744194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,8192,0.14229333400726318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,10240,0.06957955492867364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,12288,0.23107467757331002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,51200,1.0284915500217013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,8192,0.06534755229949951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,65536,1.252010663350423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,10240,0.18804711765713164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,7168,0.12792710463205972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,6144,0.09828266832563613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,5120,0.08075466420915392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,8192,0.1471493376625909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,6144,0.045159111420313515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,4096,0.0629040002822876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,7168,0.12642578283945718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,3584,0.053103999959097974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,5120,0.03857600026660495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,6144,0.11897777186499701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,3072,0.04818488823042976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,5120,0.11339644591013591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,4096,0.07628533575269911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,3584,0.02828888760672675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,3072,0.026007999976476032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,3584,0.06602578030692206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,2048,0.03292799989382426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,2560,0.023000889354281958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,3072,0.05903289053175184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,1536,0.026135999295446608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,2560,0.04991555545065138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,2048,0.020426667398876615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,1024,0.01924711134698656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,1536,0.017649778061442904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,2048,0.04183200001716614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,768,0.015558222929636637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,1536,0.033166223102145724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,1024,0.015184000134468079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,512,0.011472889118724398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,768,0.014141332772043018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,1024,0.02533066603872511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,256,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,512,0.01309333327743742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,768,0.02139555580086178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,7168,0.05076711045371162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,512,0.017263111140992906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,128,0.0070257774657673305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,64,0.006113777971929974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,256,0.012106666962305704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,256,0.01497244503762987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,32,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,128,0.01237066669596566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,3072,4096,0.03276800115903219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,128,0.014495111174053617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,3072,2560,0.04786400000254313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,64,0.014151111245155334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,3072,32,0.014524444937705994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,16384,0.20739732848273384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,65536,0.3709457715352376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,51200,0.300710227754381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,51200,0.5675155321756998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,16384,0.09520800246132745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,12288,0.16471555497911242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,16384,0.24809688991970488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,10240,0.14239466190338135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,12288,0.07256799936294556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,12288,0.1948853333791097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,51200,0.8882008658515083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,65536,1.0840293036566842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,8192,0.11001155773798625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,10240,0.06592888964547051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,7168,0.09594666957855225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,10240,0.15718932946523032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,8192,0.05370755659209358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,6144,0.08606577581829494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,7168,0.04580533173349169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,8192,0.12881510787540013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,6144,0.039855112632115684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,7168,0.11052355501386855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,4096,0.05729777945412529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,6144,0.09636622005038792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,5120,0.032837334606382586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,3584,0.04688355657789442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,5120,0.08140977885988024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,4096,0.028299556838141546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,4096,0.0656577746073405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,3584,0.025175111161337957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,65536,0.9088506698608398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,3584,0.0582559969690111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,2560,0.032053334845436945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,2048,0.026118222210142348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,3072,0.0518799987104204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,2560,0.02049066623051961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,1536,0.0214008887608846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,2560,0.044050667021009654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,2048,0.01830755505296919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,2048,0.03716177741686503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,1024,0.016211556063758004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,768,0.013191999660597907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,1536,0.01573600040541755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,1024,0.013815999858909182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,1536,0.02963288956218296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,512,0.010125333236323463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,1024,0.022729777627521094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,768,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,256,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,768,0.018944000204404194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,512,0.012357333468066322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,128,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,256,0.01163999984661738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,512,0.015508444772826301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,64,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,256,0.013392888837390475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,32,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,128,0.011380444798204633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,3072,0.04034311241573758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,128,0.012776888906955719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,64,0.01257155587275823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2560,3072,0.022807111342748005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2560,32,0.012728888955381183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,16384,0.14116533597310385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,65536,0.31698844167921275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,51200,0.40750932693481445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,65536,0.5699804623921713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,51200,0.2457973427242703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,12288,0.11513511339823405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,16384,0.08230133189095391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,10240,0.10007377465565999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2560,5120,0.06724800003899468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,16384,0.2100213368733724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,12288,0.17527911398145887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,65536,0.8941893047756619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,10240,0.056644446320003934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,51200,0.7008986473083496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,7168,0.07998755243089464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,8192,0.044080889887279935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,6144,0.07234399848514132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,8192,0.1080862217479282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,7168,0.03757955630620321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,5120,0.05497689048449198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,7168,0.09709511200586955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,6144,0.03391733434465196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,4096,0.04231199953291151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,6144,0.08437689145406087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,5120,0.028445333242416382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,5120,0.06982488764656915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,3584,0.04046400056944953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,4096,0.024159999357329473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,4096,0.05704177750481499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,3072,0.03251733382542928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,3584,0.02203377750184801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,2560,0.028320001231299505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,3584,0.05096977618005541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,3072,0.020269332660569083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,2048,0.0236799998415841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,12288,0.0626773304409451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,3072,0.045392000012927584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,2560,0.018304889400800068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,1536,0.018696000178654987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,8192,0.09256888760460748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,2560,0.039001779423819646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,1024,0.013775111072593264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,2048,0.016524444023768108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,10240,0.13793867164187962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,768,0.01109333336353302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,2048,0.03257688879966736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,1024,0.013010666602187686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,1536,0.02607911162906223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,512,0.008600888980759515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,256,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,1024,0.019871110717455547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,768,0.012089778151777057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,128,0.005802666561471091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,512,0.011358222199810876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,768,0.01696888936890496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,64,0.005032888717121548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,512,0.01387555566098955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,2048,32,0.005472888963090048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,256,0.011931555966536203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,128,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,128,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,64,0.011062222222487131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,2048,32,0.011366222467687396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,65536,0.46299288007948136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,65536,0.25378666983710396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,51200,0.34739822811550564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,16384,0.12478044297960068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,1536,0.014917333920796713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,51200,0.21243288781907824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,65536,0.6955955293443469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,16384,0.07034577925999959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,10240,0.07849422428343031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,51200,0.5415964656405979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,16384,0.176273783047994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,12288,0.05848977963129679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,8192,0.062328888310326465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,2048,256,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,10240,0.04977066649330986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,12288,0.1492640044954088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,8192,0.037230223417282104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,6144,0.04997244477272034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,7168,0.033462223079469465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,8192,0.09040621916453044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,5120,0.03964089022742377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,7168,0.07842844724655151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,6144,0.030243555704752605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,4096,0.03176888823509216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,6144,0.06786400079727173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,5120,0.02564711040920681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,3584,0.02832711074087355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,4096,0.021145777569876775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,5120,0.056883553663889565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,3072,0.024663999676704407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,4096,0.04825422167778015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,3584,0.019018666611777413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,2560,0.0224897778696484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,3072,0.017847999930381775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,3584,0.04201333390341865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,2048,0.018000889155599806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,3072,0.036933332681655884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,2560,0.015992888145976596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,1536,0.013863999810483722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,12288,0.09669333034091526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,2560,0.031459556685553655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,2048,0.01553066737122006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,1024,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,768,0.008723555339707268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,2048,0.02700622214211358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,1536,0.01313333296113544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,512,0.007143110864692264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,1024,0.011624889241324531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,1536,0.02212000058756934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,7168,0.05269955595334371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,256,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,768,0.011037333144081963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,1024,0.01684533390733931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,10240,0.11013333002726237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,128,0.0048062221871482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,512,0.01035911093155543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,768,0.01479111115137736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,64,0.0042071110672420925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,256,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1536,32,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,512,0.012065778175989786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1536,128,0.009719111025333405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,128,0.010002666877375709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,64,0.009670221971140968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,32,0.009662222531106737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,65536,0.202857772509257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,65536,0.3604026635487874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,51200,0.2903689013587104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,16384,0.09136178096135457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,51200,0.16292533609602186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,12288,0.07245422071880765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,16384,0.0554711123307546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,65536,0.5346337954203287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,16384,0.15056799517737493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,10240,0.05846666627460056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,51200,0.4269555674658881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,12288,0.04666577776273092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,8192,0.04914222160975138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,12288,0.10515644152959187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,7168,0.04261777798334757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,10240,0.04211911227968004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,8192,0.02913155489497715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,6144,0.03533688849873013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,10240,0.08794133530722724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,7168,0.026111111044883728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,8192,0.07118222448560926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,5120,0.027279110418425664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,4096,0.021306667062971327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,7168,0.06287999947865804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,6144,0.024566221568319533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1536,256,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,3584,0.020449777444203694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,5120,0.01982399986849891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,6144,0.055065777566697865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,3072,0.016576889488432143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,4096,0.01715377800994449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,5120,0.046719110674328275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,3584,0.016303999556435477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,2560,0.014546665880415173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,4096,0.03820888863669501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,3072,0.014831110835075378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,3584,0.03435644507408142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,1536,0.0101724440852801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,3072,0.030553778012593586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,2048,0.01238222254647149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,2560,0.02625777820746104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,1024,0.007805333369308048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,2048,0.02240000002914005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,1536,0.01142133358452055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,768,0.006800888727108638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,1536,0.018439110782411363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,1024,0.0103493332862854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,512,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,768,0.010031999813185798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,1024,0.014133334159851074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,256,0.004591111093759537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,768,0.012376888758606382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,512,0.009367110828558603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,128,0.0038844446341196695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,256,0.008989333278603023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,512,0.01034044474363327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,64,0.0034968890249729156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,32,0.0035013332962989807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,256,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,128,0.008463111188676622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,64,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,65536,0.23304088910420737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,1024,2048,0.0120319997270902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,51200,0.1798248953289456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,65536,0.20089511076609293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,2560,0.013825777504179211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,16384,0.07319821914037068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,12288,0.054357333315743335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,65536,0.46028709411621094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,51200,0.15892710950639513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,16384,0.054914666546715625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,10240,0.04196444484922621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,51200,0.3592488765716553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,16384,0.11809778213500977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,12288,0.04387733340263367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,8192,0.03036266565322876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,7168,0.029230223761664495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,12288,0.09274755583869086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,8192,0.02781244450145298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,6144,0.02453688946035173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,7168,0.025394666526052687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,1024,128,0.009008888569143083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,7168,0.05390488770272997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,6144,0.022428444690174524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,1024,32,0.008159110943476358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,6144,0.04695999953481886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,5120,0.01976799964904785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,4096,0.017369776964187622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,3584,0.015427556302812366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,5120,0.039769778649012245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,4096,0.01720533271630605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,3072,0.013805333111021252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,3584,0.015915556086434256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,3584,0.02927466564708286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,3072,0.014781332678265043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,2560,0.011925333076053195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,3072,0.02647022240691715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,10240,0.03849333193567064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,2048,0.01036622209681405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,2560,0.013749333719412485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,10240,0.07560000154707167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,2560,0.02272355556488037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,1536,0.00868888861603207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,8192,0.06112800041834513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,2048,0.012432888978057437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,1024,0.00674933319290479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,5120,0.021459556288189355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,1536,0.011417778001891242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,768,0.005848888721730974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,2048,0.019575999842749704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,512,0.004822222308980094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,1024,0.01033688916100396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,1536,0.01609866652223799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,768,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,256,0.0041751112374994485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,1024,0.012709333664841123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,512,0.009322666459613377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,128,0.003572444534964032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,768,0.011417778001891242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,256,0.009019555317031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,512,0.009335999687512716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,32,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,4096,0.03307644526163737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,256,0.007831999825106727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,128,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,64,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,65536,0.20279467105865479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,768,32,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,51200,0.1593493355645074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,65536,0.15464622444576687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,16384,0.05041244294908312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,51200,0.12338488631778294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,65536,0.402021328608195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,12288,0.03881333271662394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,16384,0.04306844539112515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,10240,0.03187822302182516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,16384,0.1054426696565416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,12288,0.03452355662981669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,8192,0.02386933399571313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,10240,0.030653334326214258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,10240,0.06742133034600152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,768,64,0.0033760000434186724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,7168,0.020949333906173706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,8192,0.021401777863502502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,768,128,0.008957333034939235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,6144,0.018951111369662814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,7168,0.019648889700571697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,6144,0.017398221625222098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,5120,0.01628177695804172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,7168,0.048271109660466514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,6144,0.04219733344184028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,4096,0.013331555657916598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,5120,0.015437333120240105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,3584,0.01200177768866221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,4096,0.013455111119482251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,5120,0.03599111239115397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,3072,0.010547555155224271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,3584,0.012778666284349231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,4096,0.02957333458794488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,2560,0.009394666386975182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,3584,0.026724444495307073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,2048,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,3072,0.012042666474978128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,2560,0.011568889021873474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,3072,0.023722666833135817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,1536,0.006809777683681912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,2048,0.010715555813577441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,2560,0.020679111282030743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,12288,0.08039822181065877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,2048,0.017825777331988018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,768,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,1536,0.014797333214018079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,1024,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,8192,0.05448266532686022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,512,0.003839111162556542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,1024,0.011531555818186866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,768,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,256,0.0032319999817344877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,768,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,128,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,256,0.008160889148712158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,512,0.008616000413894653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,64,0.00254133323000537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,256,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,32,0.003020444470975134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,128,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,128,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,65536,0.11664444870418972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,64,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,32,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,512,51200,0.3190506564246284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,51200,0.09485244750976562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,16384,0.03712177938885159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,65536,0.15289689434899226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,512,1024,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,51200,0.12211910883585612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,1536,0.010012444522645738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,12288,0.028265777561399672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,16384,0.04288622074657016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,51200,0.2561075687408447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,16384,0.08614666594399346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,10240,0.021178666088316176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,8192,0.01696622206105126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,12288,0.0350835555129581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,12288,0.06481955448786418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,512,512,0.008625778059164682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,10240,0.03087110983000861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,7168,0.015453333655993143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,6144,0.013112000293201871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,8192,0.02083555526203579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,10240,0.05469155642721388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,7168,0.018561777141359117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,8192,0.04440088735686409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,5120,0.011723555624485016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,4096,0.009920888476901585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,6144,0.01742844449149238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,7168,0.03951199849446615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,3584,0.008048888709810045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,5120,0.01533600025706821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,6144,0.034754667017194964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,3072,0.007705777883529663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,4096,0.01422755585776435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,5120,0.02962933315171136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,2560,0.006908444480763541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,4096,0.024812445044517517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,3072,0.012076444096035428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,2048,0.005992888990375731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,3584,0.021998221675554912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,2560,0.011493333511882357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,3072,0.019779556327395968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,1536,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,2560,0.017232000827789307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,65536,0.3247048854827881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,2048,0.010695111420419482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,1024,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,768,0.0041804446114434134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,1536,0.010110222631030613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,2048,0.014825777875052558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,512,0.0035413333939181436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,1024,0.009308444129096137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,1536,0.012441777520709567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,256,0.0028355556229750314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,768,0.008977777428097194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,1024,0.010072888599501716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,128,0.002529777793420686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,512,0.008512000242869059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,768,0.00868977771864997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,64,0.00240622212489446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,256,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,512,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,256,32,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,256,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,128,0.008018666671382057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,65536,0.09066222111384074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,128,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,51200,0.07300533188713922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,64,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,16384,0.028311110205120508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,256,32,0.005786666853560342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,65536,0.15556355317433676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,12288,0.02222844461599986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,51200,0.1234275499979655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,256,3584,0.012787555654843649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,16384,0.0437084436416626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,65536,0.2842613326178657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,10240,0.017875555488798354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,8192,0.013190222283204397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,51200,0.22270311249627006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,16384,0.07475466860665216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,7168,0.011781333221329583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,10240,0.03115733464558919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,12288,0.056814220216539174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,6144,0.010381333529949188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,8192,0.021977777282396953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,10240,0.04798755711979336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,5120,0.009175111022260454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,7168,0.02068000038464864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,8192,0.039043555657068886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,4096,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,7168,0.03463644451565213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,6144,0.01793155570824941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,3584,0.007149332927332983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,5120,0.016232000456915963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,6144,0.030299554268519085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,3072,0.0063288890653186375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,4096,0.014138667119873894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,2560,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,3584,0.013180444637934366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,4096,0.021881777379247878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,2048,0.004848000076082018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,3072,0.012531555361217923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,3584,0.019888000355826486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,1536,0.004566222015354368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,2560,0.011767110890812345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,2048,0.010967999696731567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,2560,0.015785778562227886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,1024,0.0035297779573334586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,2048,0.01367200000418557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,1536,0.010281778044170803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,1536,0.011419555379284753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,768,0.0032035555276605818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,1024,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,512,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,12288,0.03548889027701484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,768,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,256,0.0025271111064487034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,768,0.008319999608728621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,512,0.00831200016869439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,128,0.0024604443460702896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,512,0.007228444847795699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,256,0.00812533332241906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,256,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,64,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,128,32,0.0024604443460702896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3841,128,128,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,128,0.005391110976537068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,5120,0.02609688871436649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,65536,0.08909244669808282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,64,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,51200,0.07141155666775174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,32,0.005351999981535806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,16384,0.024699555502997503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,3072,0.017765333255132038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,12288,0.017295110556814406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,10240,0.011809777882364062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,8192,0.011419555379284753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,65536,0.2856169011857775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,16384,0.07437777519226074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,12288,0.056849777698516846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,7168,0.010062221851613786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,10240,0.047967112726635404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,6144,0.009445333646403419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,5120,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,4096,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,8192,0.03877066572507223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,3584,0.006308444258239534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,3072,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,7168,0.03432622220781114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,2560,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,5120,0.02581688927279578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,4096,0.021670222282409668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,2048,0.004536889079544279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,1536,0.0038906666967603895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,3072,0.017687110437287223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,1024,0.0032248888164758682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,2560,0.015573332707087198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,768,0.002934222213096089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,2048,0.013560000393125745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,512,0.0028488888508743713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,1536,0.011387555963463254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,256,0.0024604443460702896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,1024,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,128,0.0021626665774318907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,768,0.008072888685597314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,64,0.0018968888454967076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,512,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,64,32,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,256,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,128,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,51200,0.222417778438992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,65536,0.08638577991061741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,51200,0.06787911388609144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,128,1024,0.009375111096435124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,32,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,12288,0.015261333849694995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,6144,0.030084444416893855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,10240,0.011300444602966309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,3584,0.01960177719593048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,65536,0.2840800020429823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,16384,0.0746213330162896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,8192,0.010097777677906884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,51200,0.222799989912245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,7168,0.009266667068004608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,12288,0.0564791096581353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,6144,0.008630221916569604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,5120,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,4096,0.009754666851626502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,10240,0.047945777575174965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,8192,0.03886311252911886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,3584,0.008173333273993598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,7168,0.034266667233573064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,3072,0.007383999725182851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,6144,0.029874665869606867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,2560,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,2048,0.004534222185611725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,4096,0.02165777815712823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,1536,0.0038524443904558816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,3584,0.020145777199003432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,3072,0.01818933255142636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,1024,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,2560,0.015848888291252982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,768,0.0029031110720502008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,2048,0.013753777576817406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,512,0.002542222125662698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,1536,0.011422221859296164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,64,64,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,256,0.0024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,16384,0.02398666739463806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,1024,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,768,0.00812000036239624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,64,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,32,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,256,0.005883555445406172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,128,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,64,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,32,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,5120,0.02574311031235589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3841,32,128,0.002151111140847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3841,32,512,0.0071937781241205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,16384,1.5766337712605794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,16384,4.565236409505208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,16384,5.418111165364583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,51200,6.131014929877387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,12288,3.691237131754557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,16384,5.025919172498915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,12288,3.788781483968099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,12288,1.1663111580742729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,10240,0.9866364796956381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,10240,3.1024169921875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,12288,3.4786790211995444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,10240,3.113226572672526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,8192,0.7966382238599988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,8192,2.535973230997721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,10240,2.882918251885308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,8192,2.450968000623915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,8192,2.3644960191514754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,7168,0.6662151018778483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,51200,14.971042209201388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,7168,2.318794674343533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,7168,2.067932340833876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,6144,0.5928604337904189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,7168,2.0715821584065757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,6144,1.7282684114244249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,6144,1.8585955301920574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,5120,1.4655502107408311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,5120,0.5029653443230523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,6144,1.730575137668186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,4096,1.0930364396837022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,5120,1.6034862730238173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,4096,0.41506489117940265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,5120,1.4708489312065973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,4096,1.2127822240193684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,3584,0.36862932311164004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,3584,1.039579603407118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,3584,1.0447431140475802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,4096,1.1399208704630535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,3072,0.8811048931545682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,3072,0.33393332693311906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,3584,1.040520879957411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,2560,0.8097404374016656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,3072,0.9580044216579862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,2560,0.28614311748080784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,3072,0.8228053516811795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,2560,0.7802346547444662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,2048,0.5949671003553603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,2560,0.7170204586452908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,2048,0.23854311307271323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,1536,0.44859732521904844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,2048,0.742503113216824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,2048,0.6270124647352431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,1536,0.19312532742818198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,1024,0.3190053304036458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,1024,0.1540657811694675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,1536,0.5703511238098145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,1536,0.47484710481431747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,1024,0.3726622263590495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,768,0.2614062097337511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,1024,0.3490275541941325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,512,0.17278756035698783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,768,0.14300799369812012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,768,0.2982097731696235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,256,0.12624977694617376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,768,0.28131288952297634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,512,0.1217377715640598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,51200,22.212576972113713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,512,0.3239084349738227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,128,0.1034266683790419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,512,0.23070134056939018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,256,0.10592266586091782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,64,0.08691110875871445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,256,0.22367644309997559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,65536,128,0.10389244556427002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,256,0.18348266018761528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,128,0.22149244944254556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,128,0.1482871108584934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,64,0.21592444843716094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,65536,32,0.21763822767469618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,65536,32,0.08823466963238186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,65536,51200,17.504321628146702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,51200,4.821206410725911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,65536,6.287055969238281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,16384,4.388985951741536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,51200,11.689519246419271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,16384,4.685872819688585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,16384,1.2015386157565646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,65536,15.073316786024305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,12288,2.9086729685465493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,16384,3.889068603515625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,12288,2.963475545247396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,12288,0.9053351084391276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,51200,18.092183430989582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,12288,2.719294230143229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,10240,0.7447626855638293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,10240,2.806041717529297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,10240,2.431260426839193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,8192,1.9874807993570964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,8192,0.6055857870313857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,10240,2.2452871534559464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,65536,16.29595269097222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,8192,2.0743395487467446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,7168,0.5381760067409939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,8192,1.8211226993136937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,7168,1.6886498133341472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,65536,22.990134345160588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,6144,1.3276995552910698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,7168,1.6921030680338542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,6144,0.45828888151380753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,7168,1.5992737876044378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,6144,1.429414219326443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,5120,0.42305154270595974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,5120,1.337174203660753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,5120,1.2076764636569552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,6144,1.3852035734388564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,4096,1.017706659105089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,5120,1.1266151004367404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,4096,0.3898426691691081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,4096,0.92849424150255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,3584,0.2910462220509847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,3584,0.7883635626898872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,4096,0.8820275730556912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,3072,0.7822888692220052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,3072,0.27000265651279026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,3584,0.8275270991855197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,3584,0.7799697981940376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,2560,0.5983217557271322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,3072,0.696798218621148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,2560,0.2285697725084093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,3072,0.6521146562364366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,2048,0.19288267029656303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,2048,0.4848284191555447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,1536,0.3485111130608453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,2560,0.7723226547241211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,2560,0.5782097710503472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,2048,0.44754754172431094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,1024,0.23928443590799967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,2048,0.5603689087761773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,1536,0.18171200487348768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,1536,0.36304177178276914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,1536,0.5170915391710069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,768,0.1840355528725518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,1024,0.12446044551001655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,1024,0.32531823052300346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,768,0.11503733528984918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,1024,0.27199289533827037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,768,0.25423910882737905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,256,0.10427644517686631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,512,0.20119110743204752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,512,0.10053155819574992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,768,0.22124355369144014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,128,0.08510133292939927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,512,0.17845688925849068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,256,0.08528977632522583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,64,0.07102221912807889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,256,0.17735466692182753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,51200,128,0.08343555529912312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,128,0.16688711113399932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,32,0.07282311386532254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,128,0.10722577571868896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,256,0.14092532793680826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,64,0.17009245024787056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,51200,32,0.17148178153567842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,51200,512,0.14156711101531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,65536,1.9166363610161676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,51200,3.7787369622124563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,51200,1.4983848995632594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,65536,5.050281100802951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,16384,1.3458870781792536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,51200,4.224358028835721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,16384,1.507796393500434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,16384,0.3918604320949978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,51200,6.284469180636936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,12288,0.2939386632707384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,16384,1.249407132466634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,12288,0.9844915601942273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,65536,5.407598283555772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,65536,7.312096489800347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,12288,0.9697333441840278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,12288,0.9003022511800131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,10240,0.24868533346388075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,10240,0.8362035751342773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,8192,0.6614586512247721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,10240,0.9545546637641059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,8192,0.20023732715182832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,8192,0.6165946854485406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,10240,0.6904266675313314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,7168,0.649993790520562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,7168,0.17996622456444633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,8192,0.5853333473205566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,6144,0.4443795680999756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,7168,0.7130764325459799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,6144,0.1580586698320177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,5120,0.38225422965155703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,6144,0.5285946528116862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,6144,0.41713065571255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,5120,0.14566844039493138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,4096,0.2947857909732395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,5120,0.3889964421590169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,5120,0.35902044508192277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,4096,0.3212515513102214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,3584,0.25222399499681264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,4096,0.12407999568515354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,3584,0.10103555520375569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,3072,0.2339413430955675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,4096,0.311788452996148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,3584,0.2755528820885552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,7168,0.4835795296563043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,3584,0.25315554936726886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,2560,0.18524444103240967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,3072,0.08806755807664658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,2048,0.14714133739471436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,2560,0.07942133479648165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,3072,0.23927111095852324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,3072,0.22892533408270943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,2560,0.20103555255466035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,1536,0.12012355857425266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,2560,0.18912977642483184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,2048,0.06663199928071764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,1024,0.07993511358896892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,2048,0.16490311092800564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,1536,0.054767111937204994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,1536,0.13055022557576498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,2048,0.16052177217271593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,768,0.06192800071504381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,1536,0.12779288821750218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,1024,0.046226667033301465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,1024,0.09935910834206475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,512,0.04712177647484673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,768,0.04251555601755778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,768,0.07935200134913127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,256,0.03145066565937466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,768,0.08145422405666776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,512,0.03903111153178745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,128,0.025620443953408137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,512,0.06799644231796265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,256,0.034154666794670954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,512,0.06678844160503812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,256,0.06083910995059543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,64,0.023112888137499493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,256,0.053082668119006686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,16384,128,0.033879111210505165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,128,0.05786133474773831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,128,0.04054133428467645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,64,0.058442665470970996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,16384,32,0.05933422512478299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,16384,1024,0.09945955541398789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,16384,32,0.02368800010946062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,65536,1.3604968388875325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,51200,2.8689422607421875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,51200,1.122806231180827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,65536,3.5623385111490884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,16384,0.9614337285359701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,16384,0.3450675540500217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,16384,0.9011555777655708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,51200,4.344327290852864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,16384,1.5184639824761286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,51200,3.235787497626411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,65536,4.064274681939019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,12288,0.7791484196980795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,12288,0.22669156392415366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,65536,5.849730597601996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,10240,0.1952702204386393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,10240,0.6183493402269151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,12288,0.767215993669298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,12288,0.6361546516418457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,8192,0.5056639777289497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,8192,0.15898311138153076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,10240,0.6814835336473254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,10240,0.5616390970018175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,7168,0.3910826577080621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,8192,0.49645243750678164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,7168,0.13931200239393446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,6144,0.33496620919969344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,8192,0.451932430267334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,6144,0.12206933233473037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,7168,0.5945004357231988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,7168,0.3791422314114041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,5120,0.31462489234076607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,6144,0.32337866889105904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,6144,0.4839484426710341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,4096,0.21000356144375273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,5120,0.10432977808846368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,5120,0.31230934460957843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,4096,0.08600444263882107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,5120,0.27259643872578937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,3584,0.2021226618025038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,4096,0.3104471100701226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,3584,0.07728177971310086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,3072,0.17913599809010824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,3584,0.22861954900953504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,4096,0.2324915462070041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,3584,0.19882222016652426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,2560,0.13127556112077501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,3072,0.08218577835294935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,3072,0.19077955351935494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,2560,0.060679111215803355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,2048,0.12591555383470324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,2560,0.15503999922010633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,3072,0.17320533593495688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,1536,0.08594755331675212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,2560,0.1485475566652086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,2048,0.05407555566893684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,1024,0.06373511420355903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,2048,0.12777066230773926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,1536,0.04342755675315857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,1536,0.10138755374484593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,768,0.048672888014051646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,1536,0.10093955861197577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,1024,0.036892443895339966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,1024,0.07789155509736803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,512,0.038206222984525896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,1024,0.07907111114925809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,768,0.034212443563673235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,768,0.06229955620235867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,768,0.06580622328652276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,256,0.024544888072543677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,512,0.03166666626930237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,512,0.052879111634360425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,128,0.02274133265018463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,512,0.054851555162005954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,256,0.02810844447877672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,256,0.047263112333085805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,64,0.019038221902317472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,256,0.04281511240535312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,12288,32,0.01924977699915568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,12288,128,0.027858667903476294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,128,0.04484622346030342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,128,0.03282489048110114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,12288,2048,0.13436978393130833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,64,0.04546666807598538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,12288,32,0.04571466644605001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,65536,1.2175546222262912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,51200,2.8460489908854165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,51200,0.9292693667941623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,65536,3.137762705485026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,51200,3.6323254903157554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,51200,2.688709259033203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,16384,0.7861502435472277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,16384,0.2517920070224338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,16384,1.0935902065700955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,65536,3.45325067308214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,65536,4.593790266248915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,12288,0.19449599583943686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,12288,0.6866301960415311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,10240,0.5129084587097168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,16384,0.8774577776590983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,12288,0.5451662275526259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,12288,0.6859040260314941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,10240,0.17749067147572836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,10240,0.5274924702114528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,10240,0.4631351364983453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,8192,0.1343617836634318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,7168,0.3377048969268799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,8192,0.46933688057793516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,7168,0.11973422103457981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,8192,0.402986658944024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,7168,0.3506480058034261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,8192,0.38523022333780926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,6144,0.32851465543111164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,6144,0.10529066456688775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,5120,0.21548799673716226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,7168,0.3344693448808458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,6144,0.34727732340494794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,5120,0.09038933118184407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,4096,0.17268621921539307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,5120,0.2591208881802029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,6144,0.2779573334587945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,5120,0.2527262104882134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,4096,0.07503910859425862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,3584,0.17675199773576525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,4096,0.20730222596062553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,4096,0.19504177570343018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,3072,0.1431528859668308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,3584,0.06725155644946627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,3584,0.17817066775427925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,2560,0.12066222561730279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,3072,0.05860444572236803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,3584,0.17128445042504203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,3072,0.15909155209859213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,3072,0.1508906682332357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,2560,0.1337404383553399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,2048,0.09777066442701553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,2560,0.05273600088225471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,1536,0.07311289178000556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,2048,0.046625776423348315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,2560,0.12987200419108072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,2048,0.10825155840979682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,1024,0.05503288904825846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,1536,0.08657244178983901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,2048,0.10926755269368489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,1536,0.08820444345474243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,768,0.0439208878411187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,1024,0.03276622295379639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,1024,0.06629866361618042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,512,0.028933333026038274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,768,0.05308622121810913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,768,0.05756800042258369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,512,0.04495911134613884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,512,0.02786488996611701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,256,0.021300445000330608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,512,0.04781422350141737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,256,0.02453600035773383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,256,0.04061333338419596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,128,0.018526222970750596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,256,0.037969777981440224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,1536,0.03859644465976291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,64,0.01663377715481652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,128,0.02390044430891673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,128,0.03832177652253045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,10240,32,0.017507554756270517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,128,0.028747555282380845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,64,0.03866844375928243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,10240,1024,0.06827821996476915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,10240,768,0.030489779180950586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,10240,32,0.03902666767438253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,51200,51200,12.8551025390625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,65536,0.9661341773139106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,51200,0.75455109278361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,51200,2.1389386918809676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,16384,0.89263916015625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,65536,2.5393920474582248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,16384,0.2195084492365519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,12288,0.5186613400777181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,16384,0.6113271183437771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,16384,0.9439200295342339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,51200,3.327865812513563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,12288,0.5174381997850206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,12288,0.15615288416544595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,65536,2.7897298600938587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,12288,0.46284977595011395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,10240,0.13302400377061632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,65536,4.136122809516059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,10240,0.4429608980814616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,8192,0.3289857705434163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,10240,0.368597322040134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,8192,0.11445600456661648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,8192,0.3500844372643365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,7168,0.25964177979363334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,7168,0.09771377510494655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,8192,0.3056453333960639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,6144,0.2049431138568454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,7168,0.32895289527045357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,6144,0.08402133650249904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,7168,0.2672799958123101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,5120,0.1703582207361857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,6144,0.2570399973127577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,4096,0.14678666326734754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,10240,0.3820524480607774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,6144,0.22822043630811903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,51200,2.1880853441026473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,5120,0.20722133583492705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,3584,0.14303021960788304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,4096,0.062282668219672314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,5120,0.19292088349660239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,4096,0.1713191138373481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,3072,0.11749154991573757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,3584,0.05577244361241659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,3584,0.14482043849097356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,4096,0.16842133469051787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,2560,0.09857066472371419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,3584,0.14333866702185735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,3072,0.049714667929543384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,3072,0.12591555383470324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,2048,0.08399466673533122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,3072,0.12611822287241617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,2560,0.04462577899297079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,2560,0.10854311121834649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,1536,0.060686224036746554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,2048,0.03848888807826572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,2560,0.11082933346430461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,2048,0.08930133448706733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,1536,0.03233333428700765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,2048,0.09203910827636719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,1536,0.07103555732303195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,768,0.034466667307747736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,1536,0.07360800107320149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,1024,0.026073777013354834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,1024,0.05441155698564318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,512,0.02444800072246128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,1024,0.05675377448399862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,768,0.024041778511471216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,768,0.04372088776694404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,256,0.01777333352300856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,768,0.049103111028671265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,512,0.02161688937081231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,512,0.03556888964441087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,128,0.014239999983045789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,512,0.04013422131538391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,256,0.019447111421161227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,256,0.03090933296415541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,64,0.013016888664828407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,128,0.019516444868511625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,256,0.03209599852561951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,128,0.029882666137483384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,32,0.013887999786270989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,8192,128,0.024894222617149353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,64,0.029824889368481104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,8192,1024,0.04177066683769226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,8192,32,0.030144890149434406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,8192,5120,0.07366489039527045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,65536,0.9708764817979602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,51200,1.5913022359212239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,65536,2.0684666103786893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,16384,0.6684417724609375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,51200,0.7220453156365289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,16384,0.23354045550028482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,16384,0.6327075428432888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,12288,0.5695830980936686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,16384,0.5567253430684408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,51200,1.9705556233723958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,12288,0.1935297780566745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,51200,2.8660159640842013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,12288,0.4653342035081651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,10240,0.3250488970014784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,65536,2.5149004194471574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,12288,0.40917065408494735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,10240,0.1340808868408203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,65536,3.4518534342447915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,8192,0.3141324520111084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,7168,0.24572711520724824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,8192,0.09999644756317139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,10240,0.3881431155734592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,10240,0.3402720027499729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,8192,0.3438453409406874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,7168,0.29729244444105357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,7168,0.09224000242021348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,6144,0.2076328860388862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,8192,0.29807999398973256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,5120,0.16477599408891466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,6144,0.21817333168453643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,5120,0.07110222180684407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,7168,0.29209065437316895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,4096,0.13170133696662054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,6144,0.21211200290256074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,5120,0.18261155817243788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,5120,0.18170666694641113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,3584,0.12588355276319715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,4096,0.06398933463626437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,3072,0.09759822156694199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,4096,0.14583822091420492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,3584,0.052468445565965444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,4096,0.1479920016394721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,3584,0.12889156076643202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,2560,0.08348355690638225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,3584,0.13007822301652697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,3072,0.11659377151065403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,2048,0.07310577895906237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,2560,0.040347556273142494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,3072,0.1239031155904134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,2560,0.0978871120346917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,1536,0.05200533403290642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,2560,0.09911200073030259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,2048,0.03411111235618591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,2048,0.08276266521877713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,1024,0.03921422362327576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,2048,0.08433600266774495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,6144,0.07824177874459161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,1536,0.028536889288160536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,1536,0.06377689043680827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,768,0.02774133284886678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,1024,0.024283554818895128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,1024,0.04841511117087471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,512,0.020999999509917364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,768,0.022146667043368023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,1024,0.0530373321639167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,768,0.039697776238123574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,768,0.04502488838301765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,512,0.01961688862906562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,512,0.031918221049838595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,128,0.014581332604090372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,512,0.037696000602510236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,256,0.01772444446881612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,3072,0.045824001232783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,64,0.012034667034943899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,256,0.03001777662171258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,7168,128,0.01794755624400245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,32,0.012337778177526263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,128,0.02664622167746226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,128,0.022728888524903193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,64,0.026458667384253606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,32,0.02641955514748891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,7168,1536,0.06893777847290039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,7168,256,0.015488000379668342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,7168,256,0.02757688860098521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,65536,0.7574142350090874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,51200,1.4731182522243922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,51200,0.5986791186862522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,65536,1.988205379909939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,16384,0.4996239874098036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,16384,0.1690151161617703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,16384,0.487960868411594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,16384,0.6144986682467991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,51200,2.341963450113932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,65536,2.185257805718316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,65536,2.971712960137261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,51200,1.7256186803181965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,12288,0.3991244369082981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,12288,0.13028532928890652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,10240,0.29855643378363717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,10240,0.1127057737774319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,12288,0.4270995457967122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,8192,0.24472533331976998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,12288,0.3598204453786214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,10240,0.34332089953952366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,10240,0.3240391148461236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,7168,0.20448977417416045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,8192,0.08718755510118271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,8192,0.29422932200961643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,7168,0.07875555753707886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,6144,0.18312356207105848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,8192,0.2437680032518175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,7168,0.2319564554426405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,5120,0.15034488836924234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,6144,0.07112889157401191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,7168,0.21248355176713732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,4096,0.1123431126276652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,6144,0.2371759944491916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,6144,0.1825662189059787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,5120,0.06043110953436958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,5120,0.16714044411977133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,3584,0.09327466620339288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,4096,0.05289777782228258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,3072,0.08134400182300143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,4096,0.1301155620151096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,5120,0.15877421696980795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,3584,0.04565599891874525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,4096,0.12955110602908665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,2560,0.07714044385486178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,3584,0.1133155557844374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,3072,0.040624890062544085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,3584,0.11732444498274062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,2048,0.06278489033381145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,3072,0.10222044255998398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,2560,0.034993777672449745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,2560,0.08619733651479085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,1536,0.04615111152331034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,3072,0.10313866535822551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,2048,0.030471109681659277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,1024,0.03124177787039015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,2560,0.08902666966120402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,1536,0.02552622225549486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,2048,0.0739333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,768,0.025314667158656653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,1536,0.05571022298600939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,1024,0.021533333592944678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,1536,0.06093155675464206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,1024,0.042816887299219765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,1024,0.048115554783079356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,768,0.019501333435376484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,256,0.013289777768982781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,768,0.035261332988739014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,768,0.0403733319706387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,512,0.01759999990463257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,512,0.028351998991436426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,128,0.010648888846238455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,512,0.033286223808924355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,256,0.024343111448817786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,64,0.009697777529557547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,256,0.025916443930731878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,32,0.01001422190003925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,128,0.01645777788427141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,128,0.023352000448438857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,6144,128,0.02004888819323646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,64,0.023369777533743117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,32,0.02314666741424137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,6144,2048,0.07106488943099976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,6144,512,0.017643555998802185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,65536,0.6594168874952528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,6144,256,0.016639999217457242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,51200,1.249408933851454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,65536,1.6113902197943792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,16384,0.37930578655666775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,16384,0.15146311124165854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,51200,0.5883768929375542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,12288,0.3060808976491292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,16384,0.47046486536661786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,51200,1.8549359639485676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,16384,0.45867734485202366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,12288,0.3887964354621039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,65536,1.9241502549913194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,51200,1.472919993930393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,10240,0.2630702124701606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,65536,2.7126693725585938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,10240,0.10497066709730361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,8192,0.2152142259809706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,10240,0.2744764486948649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,12288,0.33112888866000706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,7168,0.1827688879436917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,8192,0.24830399619208443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,10240,0.27096888754102916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,8192,0.23901245329115126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,6144,0.16037066777547201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,7168,0.19309866428375244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,6144,0.060938669575585254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,7168,0.19203466839260527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,6144,0.16564355956183538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,5120,0.12611822287241617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,12288,0.11386044820149739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,6144,0.16890845033857557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,4096,0.10735644234551324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,5120,0.13728177547454834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,5120,0.1440160009596083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,4096,0.11045777797698975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,4096,0.04403733213742574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,8192,0.08568266365263198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,4096,0.11756711535983616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,3072,0.07687911060121325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,3584,0.09762310981750488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,7168,0.07462310791015625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,3584,0.10460355546739365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,2560,0.06332177586025663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,3072,0.03518844313091702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,3072,0.0853031078974406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,3072,0.09219466977649265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,2560,0.0303457776705424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,2048,0.05340622199906243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,2560,0.07343644566006131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,5120,0.05269955595334371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,2560,0.08045333623886108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,2048,0.026438222991095647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,2048,0.060679111215803355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,3584,0.09070399734708999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,1024,0.026352889007992212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,1536,0.022631999519136217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,1536,0.048787554105122886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,768,0.020114666885799833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,1536,0.05550933215353224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,1024,0.018684445155991446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,1024,0.03696088989575704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,3584,0.039270222187042236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,512,0.016102222932709586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,1024,0.04327288932270474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,768,0.017303999927308824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,768,0.03074399961365594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,256,0.0124177775449223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,768,0.036836445331573486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,512,0.01649599936273363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,128,0.009925333162148794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,512,0.024472888973024156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,512,0.03074222140842014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,256,0.02111022174358368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,256,0.023056889573733013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,64,0.0107342220014996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,32,0.011364444262451597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,128,0.020446222689416673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,5120,1536,0.040228443013297185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,128,0.018216888109842937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,64,0.02049688829316033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,5120,32,0.020247111717859905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,5120,2048,0.0676106678114997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,65536,0.5320533116658529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,256,0.015065777632925244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,5120,128,0.01481244464715322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,65536,1.2998133765326607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,51200,0.9804070790608724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,51200,0.41945777999030215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,16384,0.35199999809265137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,16384,0.3710800011952718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,51200,1.515610694885254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,16384,0.12507555219862196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,12288,0.281622224383884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,65536,1.8789359198676214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,65536,1.5694248411390517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,12288,0.09781066576639812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,10240,0.22458489735921225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,12288,0.26753332879808217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,10240,0.08033777607811822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,8192,0.1729582283231947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,10240,0.23444000879923502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,12288,0.2746231026119656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,51200,1.1446800231933594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,8192,0.06839644246631198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,7168,0.15926044517093235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,10240,0.27239378293355304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,8192,0.17894221676720512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,6144,0.12316889233059353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,7168,0.05960355864630806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,8192,0.18706577354007295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,7168,0.15902754995557997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,5120,0.11683289210001628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,7168,0.19411022133297393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,6144,0.05168888966242472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,6144,0.13746489418877497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,4096,0.09347377883063422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,6144,0.14106310738457575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,16384,0.36658310890197754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,5120,0.04367288947105408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,5120,0.11408800548977321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,3584,0.07838488949669732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,5120,0.12045689423878987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,4096,0.03709777858522203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,4096,0.09310399823718601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,3072,0.06817511055204603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,4096,0.10034577714072333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,3584,0.032580445210138954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,3584,0.08186399936676025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,2560,0.05307288964589437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,3072,0.029176000091764662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,3584,0.09012711048126221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,3072,0.07279111279381646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,2048,0.04615022076500786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,2560,0.02640977833006117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,3072,0.07869422435760498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,1536,0.034632888105180525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,2560,0.06168888674841987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,2048,0.02264533274703556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,2560,0.06815022230148315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,1024,0.024754666619830664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,2048,0.051706665092044406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,1536,0.018988443745507132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,2048,0.0583911140759786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,768,0.018233777748213876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,1536,0.04096711013052199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,1024,0.016588444511095684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,1536,0.04826400015089247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,512,0.015229332778188916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,1024,0.03082044588194953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,768,0.015164444843928019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,1024,0.036828445063696966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,256,0.011733333269755045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,768,0.026327111654811438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,512,0.014139556222491793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,768,0.031158222092522517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,512,0.02072888943884108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,128,0.007786666353543599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,512,0.025747555825445387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,256,0.013472889032628802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,64,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,256,0.018009778526094224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,4096,128,0.013267555170589022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,256,0.020797333783573575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,128,0.017274666163656447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,4096,128,0.017248888810475666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,64,0.017314665847354464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,4096,32,0.017214222086800467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,65536,0.4490666919284397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,65536,1.0365271038479276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,51200,0.8092835744222006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,4096,32,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,16384,0.24440267350938585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,51200,0.36330578062269425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,16384,0.10934488640891181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,16384,0.35063732994927305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,51200,1.252279069688585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,65536,1.6533955468071833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,12288,0.21947377257876924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,16384,0.32178934415181476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,65536,1.3868719736735027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,12288,0.0861004458533393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,10240,0.18201777670118544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,12288,0.23800977071126303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,8192,0.15514222780863443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,10240,0.08218844731648763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,12288,0.25914043850368923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,7168,0.12376444869571263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,8192,0.058357331487867564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,10240,0.2596071031358507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,10240,0.20425244172414145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,51200,1.0509982638888888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,6144,0.10875555541780259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,7168,0.05309066507551405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,8192,0.16569244861602783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,7168,0.14191199673546687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,5120,0.08666577604081895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,6144,0.04698933495415581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,7168,0.15249955654144287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,4096,0.07094844182332356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,6144,0.15011466874016657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,5120,0.038591110044055514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,5120,0.1041111085149977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,3584,0.06464799907472399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,4096,0.03337422344419692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,4096,0.08797422382566665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,4096,0.09173778030607437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,3072,0.04554666744338142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,3584,0.030229333374235366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,3584,0.07413333654403687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,3584,0.08059910933176677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,2560,0.04545066754023234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,3072,0.027080888549486797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,3072,0.06413777669270833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,2048,0.03578044308556451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,3072,0.07277955611546834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,2560,0.024028443627887305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,2560,0.0547928876347012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,2560,0.062237335575951465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,2048,0.020958221620983545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,1536,0.025855110751258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,2048,0.04643822378582425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,6144,0.12145955032772487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,2048,0.053583999474843345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,1024,0.019149333238601685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,1536,0.0181688881582684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,1536,0.03710844450526767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,768,0.014673777752452426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,1536,0.0428017775217692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,5120,0.10914933019214207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,1024,0.015591111448076038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,1024,0.02804888950453864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,512,0.01146577795346578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,1024,0.034295999341540866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,768,0.014547554983033074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,256,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,768,0.02345688806639777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,512,0.013420444395807056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,768,0.028916445043351915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,512,0.018622222873899672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,128,0.007370666497283512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,512,0.02381333377626207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,256,0.012431110772821637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,64,0.006743999818960826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,256,0.016336000627941556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3584,32,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,8192,0.1820488903257582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3584,128,0.012439111040698158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,128,0.015846222639083862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,128,0.016563556260532804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,64,0.015469332536061605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3584,32,0.016117332710160148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,65536,0.42396977212693954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,65536,0.8810950915018717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,51200,0.7311200035942925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3584,256,0.019998222589492798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,16384,0.27007023493448895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,51200,0.3322853247324626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,65536,1.3633421791924372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,51200,1.0348773532443576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,16384,0.10434666607115005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,65536,1.1697599622938368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,16384,0.2819777859581841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,12288,0.21045866277482775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,10240,0.1688079966439141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,12288,0.07972533173031278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,16384,0.2827422353956434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,12288,0.2230302227867974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,8192,0.1385093397564358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,51200,0.9296817779541016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,12288,0.23420800103081596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,10240,0.07385066482755873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,10240,0.1778764459821913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,7168,0.11842044194539388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,10240,0.1964684459898207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,7168,0.04609422220124138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,8192,0.141975998878479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,6144,0.1013066636191474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,7168,0.12711289193895128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,8192,0.15070311228434244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,5120,0.08418489164776272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,6144,0.040758222341537476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,7168,0.14999200238121882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,4096,0.06232533189985487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,6144,0.10914755529827541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,5120,0.03899111019240485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,6144,0.11435555087195502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,3584,0.0576488905482822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,5120,0.09306044711007012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,4096,0.029211554262373183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,5120,0.09945866796705459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,4096,0.0760657787322998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,3072,0.048004445102479726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,3584,0.02612977723280589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,4096,0.08185244268841214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,2560,0.04236711064974467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,3584,0.06559377908706665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,3072,0.023577777875794306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,3584,0.07497155666351318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,3072,0.05722044573889839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,2048,0.03369777732425266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,2560,0.02156177825397915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,3072,0.06563911173078749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,1536,0.023053333163261414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,2560,0.04985688792334663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,2048,0.018738667170206707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,1024,0.015971556305885315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,2560,0.05656088723076714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,8192,0.054991109503640064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,2048,0.04208177659246656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,1536,0.016673778494199116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,768,0.013228444589508904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,2048,0.04729333188798693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,1536,0.033334222104814314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,1024,0.014395554860432943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,512,0.011053333679835001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,1536,0.038763556215498186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,768,0.013394667042626275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,1024,0.025040000677108765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,256,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,1024,0.03022222386466132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,768,0.02113866640461816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,512,0.01258133351802826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,768,0.026245334082179602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,512,0.01716977854569753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,128,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,512,0.021640888518757288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,256,0.011775111158688864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,64,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,3072,32,0.006350222147173352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,3072,128,0.011709333293967776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,256,0.014875556031862894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,256,0.019658666517999437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,128,0.014153778553009033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,3072,128,0.015396444333924187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,64,0.014153778553009033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,3072,32,0.014383110735151501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,65536,0.36747466193305117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,51200,0.6373662418789333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,65536,0.9093732833862305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,16384,0.23037155469258627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,51200,0.2930408848656548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,51200,0.8281120194329156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,65536,1.1096009148491752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,16384,0.09596177604463364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,12288,0.16305777761671278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,51200,0.8686399989657931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,65536,1.1179315778944228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,10240,0.13480177190568712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,12288,0.07345422108968098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,12288,0.19465333885616729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,16384,0.38506044281853574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,8192,0.11273954974280463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,10240,0.06235111422008938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,12288,0.21899021996392143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,10240,0.16136088636186388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,7168,0.09619822104771932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,8192,0.049441777997546725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,10240,0.1925893359714084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,8192,0.1285582251018948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,6144,0.08305422465006511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,7168,0.044176889790429004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,7168,0.11447021696302627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,8192,0.14882666534847683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,5120,0.06472444534301758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,6144,0.03745955559942458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,6144,0.09695288870069717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,7168,0.12893866168128118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,4096,0.048355556196636625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,5120,0.032157331705093384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,6144,0.11282489034864639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,5120,0.08134133285946317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,16384,0.2835555606418186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,4096,0.02736088964674208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,5120,0.09561155239741008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,3072,0.03782755467626784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,4096,0.06548088788986206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,3584,0.024832000335057575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,4096,0.0788559979862637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,3584,0.0575795571009318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,2560,0.031572444571389094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,3072,0.022392888863881428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,3584,0.07125155793295966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,2048,0.025807110799683466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,3072,0.05079022381040785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,2560,0.02014755540423923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,3072,0.06353777647018433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,1536,0.02068800065252516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,2560,0.04401510953903198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,2048,0.017841777867741056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,2560,0.05459644397099813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,1024,0.014352000421947904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,2048,0.03712711069318984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,1536,0.015628443823920358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,2048,0.046558221181233726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,768,0.011865778101815118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,1024,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,1536,0.03752266698413425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,1024,0.022639110684394836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,1024,0.029155555698606703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,768,0.01278133359220293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,768,0.019256888164414298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,256,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,3584,0.04575111137496101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,768,0.02537511123551263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,512,0.012143111063374413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,512,0.015513777732849121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,128,0.006090666684839461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,256,0.011368888947698804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,512,0.021114667256673176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,256,0.01349066694577535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,64,0.005525333185990651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,32,0.005532444351249271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2560,128,0.011600000162919363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,256,0.017858665850427415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,128,0.012778666284349231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2560,128,0.013934221532609729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,64,0.012695111334323883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,32,0.01278133359220293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2560,1536,0.029622223642137315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,65536,0.5764675670199925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,65536,0.31057511435614693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2560,512,0.009383999639087254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,51200,0.4663502375284831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,16384,0.17849689059787324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,51200,0.2453413274553087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,65536,0.8906773461235894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,12288,0.12449244658152263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,16384,0.21417778068118623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,16384,0.08016622066497803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,65536,0.9714480506049262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,51200,0.6724159982469348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,16384,0.25027110841539174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,12288,0.0621982216835022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,10240,0.10989422268337673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,8192,0.08652266528871323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,12288,0.16494311226738825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,51200,0.7536666658189562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,10240,0.054133332437939115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,12288,0.18788444995880127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,7168,0.08102222283681233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,8192,0.04191644324196709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,10240,0.15926222006479898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,8192,0.10910844140582615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,6144,0.06753689050674438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,7168,0.03823022378815545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,7168,0.09352533022562663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,7168,0.11350666152106391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,6144,0.03332533439000448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,6144,0.08145599895053439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,6144,0.09807199902004665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,4096,0.04268266757329305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,5120,0.028232001596026953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,5120,0.06877244181103177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,3584,0.0406311121251848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,5120,0.08477689160241021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,4096,0.024225778049892847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,4096,0.05506489011976454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,4096,0.07049333386951022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,3584,0.021931555536058214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,3072,0.03533422284656101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,3584,0.04931022061241997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,10240,0.13093600008222792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,2560,0.028788444068696763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,3584,0.06304266717698839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,3072,0.02017866737312741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,2048,0.02349244389269087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,3072,0.04354222284422981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,2560,0.0377422206931644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,3072,0.05587289068433973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,1536,0.019308444526460435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,2560,0.04891644583808052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,5120,0.05439644389682346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,2048,0.016574222180578444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,2048,0.03177333209249709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,1024,0.013853333062595792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,1536,0.01479644411140018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,2048,0.04129244552718268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,768,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,1536,0.03422666589419047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,1024,0.012764444781674279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,1024,0.019475556082195707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,512,0.008489777644475302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,1024,0.026555554734336004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,768,0.012096000214417776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,768,0.022847111026446026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,512,0.011178666518794166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,512,0.0136808885468377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,256,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,512,0.01924888955222236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,128,0.005982222242487802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,256,0.010676444404655032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,2560,0.018240888913472492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,256,0.011702222128709158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,8192,0.12984800338745117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,64,0.005990222096443176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,256,0.016597333881590102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,2048,32,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,2048,128,0.010689777632554373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,128,0.011425777441925473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,2048,128,0.01273333364062839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,64,0.011063999599880643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,32,0.011219555305110084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,1536,0.026007110873858135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,65536,0.45807822545369464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,65536,0.2549075550503201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,2048,768,0.016522667474216886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,51200,0.3386568758222792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,16384,0.12944089041815862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,51200,0.20215466287400988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,65536,0.6854844623141818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,16384,0.07167110840479533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,12288,0.09243733353084987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,51200,0.544689761267768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,16384,0.17247377501593697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,10240,0.08304622438218859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,16384,0.20820977952745226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,65536,0.8060951232910156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,12288,0.05687555339601305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,51200,0.6239191161261665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,12288,0.1328915490044488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,10240,0.04826133449872335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,10240,0.10958666933907403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,12288,0.16260266304016113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,7168,0.05610399776034885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,8192,0.03536977701716953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,10240,0.1351084444257948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,8192,0.08906577693091498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,6144,0.04893155561553108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,8192,0.11128000418345134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,7168,0.031769778993394636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,7168,0.07761333386103313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,5120,0.040743112564086914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,7168,0.09811555676990086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,6144,0.028107557031843398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,6144,0.06807911396026611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,4096,0.029209779368506536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,6144,0.08465688758426243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,5120,0.024110222856203716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,5120,0.056468447049458824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,3584,0.027107555005285475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,5120,0.07271733548906115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,4096,0.020894222789340552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,4096,0.04660266637802124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,3072,0.022520888182852004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,4096,0.061343113581339516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,3584,0.018735110759735107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,3584,0.04184266593721178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,2560,0.019090667366981506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,8192,0.06512800190183851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,3584,0.054654220740000405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,3072,0.01734755602147844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,2048,0.016193777322769165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,3072,0.03696177734269036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,3072,0.048539555735058255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,1536,0.012677333421177335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,2560,0.03228444523281521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,2048,0.014863111906581454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,2560,0.042360888587103955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,2048,0.026866666144794885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,1024,0.009542221824328104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,2048,0.035247110658221774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,1536,0.013219555219014486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,1536,0.02186044388347202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,768,0.007756444315115611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,1536,0.029139555162853662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,1024,0.011707555916574268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,512,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,1024,0.02381422287887997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,768,0.011023999916182624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,768,0.014214222629865011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,256,0.004972444640265571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,768,0.02057422200838725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,512,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,512,0.011934222446547614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,128,0.004233777936961916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,512,0.018000000052981906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,256,0.009973333113723332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,64,0.004184000194072723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,256,0.010059555371602377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1536,32,0.0042071110672420925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,128,0.009891555541091496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,128,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,256,0.0155102229780621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1536,128,0.01221244451072481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1536,2560,0.015955555770132277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,64,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,32,0.009529777699046666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,65536,0.33179465929667157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,65536,0.203529781765408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,51200,0.2889982329474555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,16384,0.0932568907737732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1536,1024,0.01682222220632765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,51200,0.1642213397555881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,65536,0.5417751206292046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,12288,0.06898844242095947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,51200,0.4190391169653998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,16384,0.05565244621700711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,10240,0.05667733483844333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,65536,0.6526062223646376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,16384,0.18116088708241782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,12288,0.044043554200066455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,51200,0.5133040216233995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,8192,0.04569777846336365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,12288,0.13333955076005724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,10240,0.08842666943868001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,7168,0.03944622145758735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,10240,0.11289600531260173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,8192,0.028287112712860107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,8192,0.07189155287212796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,6144,0.034280889564090304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,8192,0.09329866700702244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,7168,0.025384889708624944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,7168,0.06269866890377469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,5120,0.028696888022952612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,6144,0.054232001304626465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,7168,0.08169955677456327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,6144,0.07121866941452026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,5120,0.019681778219011094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,16384,0.13696444034576416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,4096,0.02348533272743225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,5120,0.046068443192376025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,5120,0.06101510922114054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,4096,0.016951110627916124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,3584,0.02126311096880171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,12288,0.10498133632871841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,4096,0.03824799921777513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,10240,0.039155556095971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,3072,0.018516444497638278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,4096,0.0512586666478051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,3584,0.0158942225906584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,2560,0.016336888074874878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,3584,0.03432622220781114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,3072,0.014761777387724983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,3584,0.046042667494879834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,2048,0.013820444544156393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,3072,0.030468444029490154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,2560,0.013718222578366598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,3072,0.040469331873787776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,1536,0.01072177787621816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,2560,0.035417778624428645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,6144,0.022208889325459797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,2048,0.012434666355450949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,2048,0.022338666849666174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,1024,0.008375110725561777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,2048,0.030096888542175293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,1536,0.011368888947698804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,1536,0.018383110562960308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,768,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,1024,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,1536,0.02490844494766659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,1024,0.014138667119873894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,512,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,1024,0.02032088902261522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,768,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,256,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,768,0.012079999678664737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,768,0.017792888813548617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,512,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,512,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,256,0.008973333570692275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,512,0.015508444772826301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,256,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,64,0.0038480001191298165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,32,0.0039057777159743835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,256,0.013437333206335703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,128,0.008012444608741337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,1024,128,0.010747555229398938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,64,0.008017777568764156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,65536,0.2274115615420871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,2560,0.026196445027987163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,51200,0.1857066684299045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,65536,0.19890311029222277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,65536,0.45546579360961914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,16384,0.0645973351266649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,51200,0.3631760014428033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,51200,0.1581617726220025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,1024,128,0.0041742221348815495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,65536,0.6456302007039388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,16384,0.054888000090916954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,12288,0.05089422067006429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,16384,0.117777771419949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,1024,128,0.008959111240175035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,10240,0.04199555516242981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,12288,0.04327555497487386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,16384,0.17132177617814806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,51200,0.5237724516126845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,8192,0.03261244297027588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,12288,0.08957778082953559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,10240,0.038203554020987615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,12288,0.13202577167087132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,1024,32,0.008019555773999956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,7168,0.027804444233576458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,10240,0.07516622543334961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,6144,0.022535110513369244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,8192,0.06195555792914497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,10240,0.11193511221143936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,7168,0.025091555383470323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,8192,0.09215555588404338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,5120,0.020218667056825426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,7168,0.05402222275733948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,6144,0.022221333450741235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,7168,0.08045955499013265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,4096,0.016294222739007738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,6144,0.04668177829848396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,5120,0.019461333751678467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,3584,0.014420444766680399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,6144,0.07031822204589844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,5120,0.03990222348107232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,5120,0.06071199973424276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,4096,0.016965332958433364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,3072,0.012823111481136747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,4096,0.032920890384250216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,3584,0.01579466627703773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,4096,0.050365332100126475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,2560,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,3584,0.029678222205903795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,3072,0.014494222071435718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,3584,0.044158223602506846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,3072,0.02624799973434872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,2048,0.009424000150627559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,2560,0.01351199961370892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,3072,0.03939822316169739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,2560,0.02277511192692651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,1536,0.007890666524569193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,2560,0.03432888785998026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,2048,0.012424888710180918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,1024,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,2048,0.019430221782790292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,8192,0.02776622275511424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,1536,0.011386666860845355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,768,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,1536,0.016210666961140104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,1536,0.024544888072543677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,1024,0.012735111018021902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,512,0.004536889079544279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,768,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,1024,0.01993333299954732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,768,0.011064888702498542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,256,0.0035457776652442086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,768,0.01753422286775377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,512,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,512,0.009380444056457942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,128,0.0032088889016045462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,512,0.015110222829712762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,256,0.00775999989774492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,256,0.013438222308953604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,64,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,128,0.008989333278603023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,128,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,768,32,0.003160888950030009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,128,0.01038755559258991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,64,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,65536,0.20183911588456896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,768,32,0.007450666692521837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,768,2048,0.029556443293889362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,1024,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,51200,0.16057244936625162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,65536,0.1534551117155287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,16384,0.05239911211861504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,65536,0.401417785220676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,51200,0.12290755907694499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,51200,0.31595643361409503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,12288,0.038571556409200035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,768,256,0.008997333546479544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,16384,0.04303555687268575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,16384,0.10570666525099014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,65536,0.5001164542304145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,10240,0.03239999877081977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,51200,0.3962862226698134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,16384,0.13682577345106336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,12288,0.03432177835040622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,8192,0.02441155579355028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,10240,0.029770665698581274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,12288,0.08011555671691895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,7168,0.02117777698569828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,10240,0.06741422414779663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,12288,0.10587822066413032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,8192,0.02113955550723606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,6144,0.018447111050287884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,10240,0.08930133448706733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,8192,0.05424088901943631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,5120,0.01587377819750044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,7168,0.018990221950742934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,7168,0.04807910985416836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,8192,0.07277866866853502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,6144,0.017082666357358296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,4096,0.013608888619475894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,7168,0.06411466995875041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,6144,0.042057779100206166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,5120,0.015184000134468079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,3584,0.011956444217099084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,6144,0.05673155519697401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,5120,0.035939554373423256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,3072,0.011040888726711273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,5120,0.048319111267725624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,4096,0.029535998900731403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,3584,0.012472000386979846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,4096,0.040821333726247154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,2560,0.009339555270142024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,3584,0.026571555270089045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,3072,0.012078222301271228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,2048,0.008261333737108443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,3072,0.02372444503837162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,2560,0.011370666325092316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,3072,0.032256888018714056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,2560,0.020870221985710993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,2560,0.028445333242416382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,2048,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,2048,0.017761778500345018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,1024,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,2048,0.024331554770469666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,1536,0.010050666828950247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,1536,0.01482222146458096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,768,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,1024,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,1536,0.020573332905769348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,1024,0.011693333586057028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,1024,0.017190222938855488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,768,0.00906666699382994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,512,0.003911111089918348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,768,0.010061333576838175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,4096,0.013429332938459186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,256,0.0034835553831524322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,768,0.01516711049609714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,512,0.008350222474998897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,512,0.00903200027015474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,128,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,3584,0.03577422102292379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,256,0.008157333566082848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,256,0.007096000015735626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,64,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,256,0.011711999773979187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,512,128,0.00815822184085846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,32,0.0032808888289663526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,128,0.006804444309737947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,128,0.009000889129108852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,65536,0.1079839997821384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,64,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,512,32,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,51200,0.09521600272920396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,65536,0.15116000175476074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,65536,0.3263466623094347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,16384,0.03668977816899618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,51200,0.12108888891008164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,51200,0.25539644559224445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,65536,0.48708354102240664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,51200,0.382567114300198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,512,512,0.013106666505336761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,16384,0.0859617789586385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,16384,0.1322960058848063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,512,1536,0.006706666615274217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,10240,0.021028444170951843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,12288,0.06492533286412557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,12288,0.10240444209840562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,8192,0.016424889365832012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,10240,0.029339555237028334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,10240,0.055213332176208496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,7168,0.014614222778214348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,10240,0.08655644125408596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,8192,0.020298666424221463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,8192,0.0440995560752021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,6144,0.013215111361609565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,7168,0.018585777944988675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,7168,0.039488888449139066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,7168,0.062308443917168506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,5120,0.011366222467687396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,6144,0.01681688924630483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,6144,0.03470933437347412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,16384,0.04240533378389147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,4096,0.009339555270142024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,5120,0.015040889382362366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,5120,0.029672887590196397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,12288,0.03396711084577773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,5120,0.046550220913357206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,3584,0.00871022211180793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,4096,0.02440799938307868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,4096,0.013365333278973898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,3072,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,4096,0.03847555650605096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,3584,0.012494222157531314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,3584,0.02211911148495144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,2560,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,3584,0.03476622369554307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,3072,0.011855111353927188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,3072,0.019671999745898776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,2048,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,8192,0.07074666685528226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,3072,0.03068444463941786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,2560,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,1536,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,2560,0.017266665895779926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,2048,0.010751999914646149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,2560,0.027373333772023518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,2048,0.014955555399258932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,1024,0.004270222038030624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,12288,0.027905778752432928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,2048,0.023436443673239812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,6144,0.054526223076714404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,1536,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,1536,0.01260444439119763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,768,0.0038533334930737815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,512,0.0031786666562159858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,1536,0.019864888654814828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,1024,0.009133333133326637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,1024,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,768,0.00886044485701455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,256,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,1024,0.016507556041081745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,768,0.008650666309727563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,128,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,512,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,256,0.008055999875068665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,512,0.01308444473478529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,256,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,64,0.0022577777918842104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,256,32,0.0024622221373849446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,256,0.011359111302428775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,128,0.008040889269775814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,128,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,128,0.008619555168681675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,65536,0.08886400196287368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,64,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,256,32,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,51200,0.07263111405902438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,65536,0.15145156118604872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,65536,0.28266043133205837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,16384,0.027822222974565294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,51200,0.2230586740705702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,51200,0.12112088998158772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,65536,0.4798426628112793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,256,512,0.008356444537639618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,16384,0.04345688886112637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,12288,0.020303999384244282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,51200,0.377691560321384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,256,768,0.014866666661368476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,10240,0.015430221954981485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,16384,0.1306640042199029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,12288,0.03447644578086005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,8192,0.012936000194814471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,12288,0.056848002804650195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,10240,0.03052355514632331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,10240,0.04817688796255323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,12288,0.10176889101664226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,10240,0.08540444241629706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,8192,0.021485333641370136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,8192,0.03907377852333917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,6144,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,8192,0.06934489144219293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,7168,0.0192275560564465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,7168,0.03467377689149644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,5120,0.008667555948098501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,7168,0.06116533279418945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,6144,0.017543110582563613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,6144,0.03046577837732103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,4096,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,6144,0.05353066656324598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,5120,0.015648888217078317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,5120,0.0258586671617296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,3584,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,16384,0.07448533508512709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,5120,0.04558755622969734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,3072,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,4096,0.021709332863489788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,3584,0.013004444539546967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,4096,0.03773333297835456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,2560,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,3072,0.012376888758606382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,3584,0.03393155667516921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,3072,0.017640888690948486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,7168,0.011401777466138205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,2048,0.004554666578769684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,3072,0.030489779180950586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,2560,0.011704000333944956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,2560,0.015857777661747403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,1536,0.004223111189073986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,2048,0.010748444332016839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,2560,0.026759111218982275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,2048,0.013767110804716745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,2048,0.023112888137499493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,1536,0.011417778001891242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,1536,0.00998933364947637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,768,0.0031848889258172777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,1536,0.01925511161486308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,1024,0.009304000271691216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,1024,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,512,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,1024,0.015903110305468243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,768,0.008979555633332994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,768,0.008450667063395182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,4096,0.013776000175211163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,256,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,512,0.008464888566070134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,512,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,128,0.0021768889079491296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,512,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,256,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,256,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,64,0.002128888956374592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,256,0.011056000159846412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,32,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3840,128,128,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,128,0.00535733335547977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,65536,0.08595644765430027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,64,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,32,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,51200,0.0659360024664137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,16384,0.02777333392037286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,12288,0.01640533407529195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,65536,0.28441956308152944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,10240,0.012453333371215396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,16384,0.07455555597941081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,51200,0.222762664159139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,8192,0.012072000238630505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,12288,0.056679109732309975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,7168,0.010891555911964841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,768,0.014350222216712104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,6144,0.009361777868535783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,10240,0.04775289032194349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,128,3584,0.019855999284320407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,8192,0.038655999634000994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,5120,0.00814400033818351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,4096,0.006948444578382704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,7168,0.03425866696569655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,3584,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,6144,0.02978844443957011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,3072,0.005761777775155173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,2560,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,5120,0.025755556093321905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,4096,0.021725333399242822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,2048,0.004555555681387584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,3584,0.019595555133289762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,1536,0.0038542221817705366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,3072,0.01758311192194621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,128,1024,0.0035173334181308746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,2560,0.015600888265503777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,768,0.0029271110478374693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,512,0.0028328889360030494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,2048,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,256,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,1536,0.011419555379284753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,128,0.002116444417172008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,768,0.008359111017651027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,64,0.0019439999014139175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,512,0.006973333656787872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,32,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,256,0.005883555445406172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,128,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,65536,0.08207822508282132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,64,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,51200,0.06600177950329252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,32,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,16384,0.02383111086156633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,12288,0.015097777048746744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,65536,0.28263733122083873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,16384,0.07414044274224176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,51200,0.22236977683173287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,10240,0.011297778122954898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,64,1024,0.003269333392381668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,12288,0.05652444230185615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,7168,0.009154666629102495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,6144,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,10240,0.04775289032194349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,8192,0.038726223839653864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,64,1024,0.009310222334331935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,5120,0.010405333505736457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,7168,0.03410222133000692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,4096,0.008943110704421997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,3584,0.008237332933478886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,6144,0.030194666650560167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,5120,0.025776000486479864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,4096,0.02162666618824005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,2560,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,3584,0.019928889142142404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,1536,0.003822222352027893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,3072,0.017836444907718234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,1024,0.0032071111102898917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,2560,0.015809777710172866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,768,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,2048,0.013748444616794586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,512,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,1536,0.011400000088744693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,256,0.0022897778285874263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,1024,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,128,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,8192,0.009908444351620143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,768,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,64,0.0018257778137922287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,512,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,32,0.001869333287080129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,256,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,128,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,64,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3840,32,32,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,3072,0.00721777809990777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3840,32,2048,0.004552888787455029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,16384,1.5104337268405492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,12288,3.211374282836914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,16384,4.400364345974393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,12288,1.1344373491075304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,16384,5.095008002387153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,12288,3.8123139275444875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,10240,0.9624622133043078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,51200,6.08786137898763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,10240,2.8779474894205728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,8192,0.8199777603149414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,8192,2.286930720011393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,10240,3.048076417711046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,7168,0.6681822141011556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,8192,2.328274620903863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,6144,0.5731689135233561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,6144,1.8235723707411025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,7168,2.0559537675645614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,7168,2.0231075286865234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,5120,0.491014215681288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,6144,1.7898932562934027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,5120,1.509874661763509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,4096,1.07987732357449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,5120,1.410496923658583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,4096,0.40225688616434735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,4096,1.156703101264106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,3584,0.35736799240112305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,3584,1.029239124721951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,3584,0.9697857962714301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,51200,13.636734856499565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,3072,0.7634311252170138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,3072,0.32483379046122235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,3072,0.8778150876363119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,2560,0.6901822090148926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,2560,0.27554310692681205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,2048,0.6351475715637207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,2048,0.24847822719150117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,1536,0.4558986557854547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,2560,0.7220035658942329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,2048,0.5886791017320421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,1536,0.1889359951019287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,1024,0.2791128953297933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,1024,0.147615114847819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,768,0.2361475626627604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,768,0.13430578178829616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,512,0.1956782208548652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,1024,0.44490400950113934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,256,0.11396888891855876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,768,0.280786673227946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,512,0.12177688545650905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,512,0.26757065455118817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,128,0.0984684427579244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,256,0.11290666792127822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,64,0.08323466777801514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,256,0.21797511312696669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,65536,128,0.10528355836868286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,65536,32,0.08719466792212592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,128,0.20419377750820586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,64,0.20811111397213408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,32,0.2114248938030667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,1536,0.5415848626030816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3840,128,128,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,16384,3.7231920030381946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,51200,4.555272844102648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,65536,51200,22.536736382378475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,65536,6.495880974663629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,16384,1.1479351255628798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,12288,0.8579484621683756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,12288,2.656944910685221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,16384,4.030265808105469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,10240,0.7286453247070312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,12288,3.12392086452908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,51200,10.71012963189019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,10240,2.3629023234049478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,8192,0.6006728808085123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,10240,2.983320871988932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,8192,2.3257645501030817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,7168,0.610362688700358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,8192,1.9125866360134547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,7168,1.729066636827257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,6144,0.4757368829515245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,7168,1.5706453323364258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,65536,14.153565300835503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,6144,1.4610586166381836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,5120,0.3798675537109375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,6144,1.3838515811496312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,5120,1.126175986395942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,4096,0.3392728964487712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,5120,1.1936204698350694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,4096,1.1084106233384874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,3584,0.9112906985812717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,3584,0.27706578042772084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,4096,0.9189635382758247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,3072,0.7470729086134168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,3584,0.7788035604688855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,3072,0.24961956342061362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,2560,0.5607173177931044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,2560,0.2502391073438856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,2048,0.4135875436994765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,2560,0.5696755515204536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,2048,0.18913155131869844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,3072,0.719455983903673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,2048,0.47981691360473633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,1536,0.1511182255215115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,1024,0.2504275639851888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,1024,0.11905777454376221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,768,0.18854933314853242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,1536,0.3737875620524089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,1024,0.2763591077592638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,512,0.13528177473280165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,256,0.09467555416954888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,768,0.10576888587739731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,768,0.25140266948276097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,512,0.18591377470228407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,128,0.08171644475724962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,51200,17.16747368706597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,256,0.08361155456966823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,64,0.08130133152008057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,256,0.1727937724855211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,32,0.08606044451395671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,128,0.0828746689690484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,128,0.1588053305943807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,64,0.16150932841830784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,51200,1536,0.3654924498664008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,32,0.16414933734469944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,51200,512,0.09531466828452216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,65536,1.8643262651231554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,51200,1.4297030766805012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,16384,0.3879724343617757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,51200,3.5546497768825955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,16384,1.423993746439616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,16384,1.2745084762573242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,12288,0.9596800274319119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,51200,65536,22.030388726128475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,12288,0.3227013217078315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,65536,4.528320736355252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,10240,0.740140438079834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,10240,0.24301067988077799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,12288,0.9162711037529839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,8192,0.6613244480556911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,8192,0.20572000079684785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,10240,0.8669404453701444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,7168,0.5543386671278212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,8192,0.6207937664455837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,7168,0.20447377363840738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,6144,0.4219786591000027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,7168,0.516034656100803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,6144,0.19228622648451063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,5120,0.34032532903883195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,5120,0.13206222322252062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,6144,0.4464595582750108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,4096,0.2707688808441162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,4096,0.10835288630591498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,5120,0.41347466574774844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,3584,0.25253332985772026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,3072,0.2205280065536499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,3584,0.09894933303197224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,4096,0.3622844484117296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,3584,0.25672711266411674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,3072,0.09505244758394028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,2560,0.07631288634406196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,3072,0.23036000463697645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,2048,0.1440888908174303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,51200,6.624425676133897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,2560,0.1968302196926541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,65536,6.994882795545791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,2048,0.06502399841944377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,1024,0.08021866612964206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,2048,0.16297688749101427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,1536,0.057032889790005155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,768,0.06531377633412679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,1024,0.044734223021401294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,1536,0.1271644433339437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,512,0.04643644558058845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,1024,0.1014817754427592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,256,0.03212089008755154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,768,0.04450666573312548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,512,0.03837066557672288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,768,0.07679111427730985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,128,0.030631999174753826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,256,0.03345600101682875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,64,0.026846221751636926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,32,0.023786667320463393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,256,0.059280885590447314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,128,0.05503111084302267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,64,0.05568355321884155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,2560,0.17657066716088188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,32,0.05699288845062256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,16384,1536,0.1215484407212999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,16384,512,0.0647546648979187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,16384,128,0.033246222469541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,51200,1.0445031060112846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,16384,0.8443466822306315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,65536,1.5123430887858074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,16384,0.957999971177843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,16384,0.3658782111273871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,51200,2.3053910997178817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,12288,0.6709315511915418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,12288,0.225911988152398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,12288,0.7022275394863553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,10240,0.5708248880174425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,65536,3.212250603569878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,8192,0.4137537744310167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,8192,0.16032177872127956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,10240,0.6259466807047526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,8192,0.45582220289442277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,7168,0.37702756457858616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,7168,0.13701955477396646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,6144,0.3474862310621474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,6144,0.12232977814144558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,7168,0.4257004525926378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,5120,0.2970115608639187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,6144,0.3375102149115668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,4096,0.2153280046251085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,10240,0.1877439949247572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,51200,4.152093251546224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,5120,0.30744179089864093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,4096,0.08700888686709934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,4096,0.25371376673380536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,3584,0.08077689011891682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,3072,0.1600337823232015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,3072,0.06833333439297147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,2560,0.1370533307393392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,3072,0.17042756080627441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,2560,0.060263997978634305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,3584,0.195835550626119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,2048,0.11597955226898193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,2560,0.15388533804151747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,1536,0.08533510896894668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,2048,0.0510391096274058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,1024,0.06167466772927178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,2048,0.1210462252298991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,1536,0.10079999764760335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,5120,0.10346577564875285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,768,0.047352890173594155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,1024,0.03581777877277798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,512,0.03566844595803155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,1024,0.0739591121673584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,768,0.03477422065205044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,3584,0.1880160040325589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,256,0.026820444398456152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,768,0.059043553140428334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,65536,6.011012607150608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,512,0.03092977735731337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,128,0.02152266601721446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,64,0.01812000075976054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,256,0.027838223510318335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,12288,32,0.01847822301917606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,128,0.02682755556371477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,256,0.044920000765058726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,128,0.04560710986455282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,64,0.043228444125917226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,32,0.04398488998413086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,12288,1536,0.04253066579500834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,12288,512,0.050360000795788236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,65536,1.2017271253797743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,51200,0.9163519541422526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,16384,0.716485341389974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,16384,0.258407990137736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,16384,0.8127306832207574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,12288,0.18705599837832979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,51200,2.299610561794705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,12288,0.6911173396640353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,65536,2.85334841410319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,10240,0.49321598476833767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,10240,0.1616951094733344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,8192,0.4619439972771539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,8192,0.14894489447275797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,10240,0.5244035720825195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,7168,0.34314576784769696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,7168,0.1299617820315891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,6144,0.3253555562761095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,51200,3.3236185709635415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,6144,0.10805244578255548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,7168,0.4425475597381592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,5120,0.2561928961012098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,6144,0.4013813336690267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,5120,0.08730222119225396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,4096,0.19530755943722197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,4096,0.07775377564960055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,5120,0.26185067494710285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,12288,0.6692328982883029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,3584,0.16724088456895617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,65536,4.21029790242513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,3072,0.12792089250352648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,4096,0.26582132445441353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,2560,0.11655199527740479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,3072,0.06364355484644572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,2048,0.1047839985953437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,3584,0.1752284500333998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,3072,0.1457644436094496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,2560,0.05804355276955498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,1536,0.08004977968004015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,2560,0.12567822138468424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,1536,0.03918933206134372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,2048,0.10995644330978394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,8192,0.3985733456081814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,768,0.04148000147607591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,1536,0.08170221911536323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,1024,0.0313715570502811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,1024,0.06265866756439209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,768,0.029530667596393164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,256,0.021871111459202234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,768,0.05043022169007195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,512,0.02637422250376807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,512,0.04251199960708618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,128,0.016741333736313712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,256,0.023778667052586872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,64,0.014474666780895658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,256,0.03812266720665826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,128,0.02292799949645996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,32,0.014852444330851236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,3584,0.06388888756434123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,128,0.035607109467188515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,64,0.03599288728502061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,10240,32,0.03587911195225186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,10240,2048,0.047482666042115956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,512,0.03194311261177063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,65536,0.8970266977945963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,16384,0.611229313744439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,51200,0.7223031255933973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,10240,1024,0.05857511361440023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,16384,0.20417243904537627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,12288,0.5934826532999674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,51200,1.9828542073567708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,16384,0.92666138543023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,12288,0.4631031354268392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,12288,0.17823378245035806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,10240,0.4174666669633653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,10240,0.1360213359196981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,65536,2.776182174682617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,8192,0.31944889492458767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,51200,2.6082835727267795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,10240,0.39546222156948513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,7168,0.28894845644632977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,7168,0.09612888760036892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,8192,0.3462417655520969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,6144,0.24038132031758627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,7168,0.30045421918233234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,5120,0.19123821788363984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,6144,0.08311733272340563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,65536,3.5119910769992404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,6144,0.257329781850179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,5120,0.07208000289069282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,4096,0.14565689033932155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,5120,0.21567200289832222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,3584,0.12880799505445692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,4096,0.06571377648247613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,3584,0.056871996985541455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,3072,0.10891555415259467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,4096,0.16221511363983154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,2560,0.0992435548040602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,3072,0.04989955491489834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,3584,0.1394008927875095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,2560,0.043172445562150746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,2048,0.08277066548665364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,3072,0.12147288852267796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,1536,0.0614008903503418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,8192,0.10932533608542548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,2560,0.10508533318837483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,1024,0.040016889572143555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,2048,0.03666311171319749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,768,0.03138844503296746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,1536,0.031312889522976346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,2048,0.0873253345489502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,512,0.023264888260099623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,1024,0.025934222671720717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,256,0.017696888910399545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,1536,0.06896799802780151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,768,0.023052444060643513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,128,0.01383733335468504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,1024,0.05152533451716105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,512,0.020766221814685397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,768,0.04195377892918057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,64,0.011710222396585675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,8192,32,0.012137778103351593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,512,0.03398666779200236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,128,0.01882577770286136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,256,0.029913776450686987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,128,0.02847466535038418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,64,0.028416888581381902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,8192,32,0.02808800008561876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,16384,0.4969644546508789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,51200,0.6849102444118924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,65536,0.9255644480387369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,8192,256,0.018992000155978732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,16384,0.28854666815863717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,16384,0.5884648958841959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,51200,1.6772968504163954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,12288,0.49832799699571395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,65536,1.9363227420383031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,12288,0.16162578264872232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,12288,0.4174648920694987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,10240,0.30778490172492134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,10240,0.14509956041971842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,8192,0.26421599917941624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,7168,0.2156906657748752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,8192,0.0991706649462382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,10240,0.38602044847276473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,6144,0.20432710647583008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,8192,0.32460888226826984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,7168,0.0946577787399292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,51200,2.3700283898247614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,7168,0.2558142344156901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,6144,0.07829688655005561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,5120,0.169887105623881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,6144,0.20805599954393175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,4096,0.135827554596795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,5120,0.07040977478027344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,4096,0.05624889002905952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,3072,0.09640355242623223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,3584,0.11898844771915013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,3584,0.05537955628501045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,5120,0.17461244265238443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,4096,0.13949155807495117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,2560,0.0883751114209493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,2048,0.07066933314005534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,2560,0.03938488827811347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,1536,0.05113510953055488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,2048,0.03355200091997782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,2560,0.10277689165539211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,2048,0.08032000064849854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,1024,0.036834667126337685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,65536,3.667327033148872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,1536,0.02814133299721612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,768,0.02869422237078349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,1024,0.028516444895002577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,1024,0.047318223449918956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,512,0.020761777957280476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,768,0.023334221707450017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,768,0.03832177652253045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,256,0.015847999188635085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,512,0.019827556278970506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,512,0.030615111192067463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,128,0.013576889203654395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,256,0.018068444397714403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,256,0.026549332671695288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,3584,0.12274132834540473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,128,0.01827111177974277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,128,0.025400888588693406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,32,0.011725333001878528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,64,0.025352888637118872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,32,0.025103999508751765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,1536,0.06092355648676554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,7168,3072,0.04814577764934964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,7168,3072,0.10794488588968913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,65536,0.7569377687242297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,16384,0.47358131408691406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,65536,1.8978640238444011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,51200,1.5774639977349176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,51200,0.5963119930691189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,16384,0.16597421964009604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,16384,0.6856320169236926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,12288,0.36342843373616535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,12288,0.14036088519626194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,7168,64,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,10240,0.290376000934177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,12288,0.4071804417504205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,10240,0.1198844379848904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,8192,0.2629511091444227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,8192,0.1031351089477539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,10240,0.3159724341498481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,7168,0.18519911501142713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,65536,2.6970382266574435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,8192,0.23734400007459852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,7168,0.08397333489523993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,6144,0.19894222418467203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,6144,0.07207644648022123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,5120,0.14002222485012478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,7168,0.22016621960533989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,4096,0.11148711045583089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,5120,0.0594008896085951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,51200,1.960577858818902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,6144,0.17997333738538954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,3584,0.10523821910222371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,4096,0.05251555641492208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,5120,0.15110666222042507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,3072,0.08765955766042073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,2560,0.06933510965771146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,3584,0.044477333625157676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,4096,0.12732711103227404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,3072,0.03919466667705112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,2048,0.060342219140794545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,1536,0.044933332337273486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,3584,0.13429510593414307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,2560,0.03729066583845351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,3072,0.09406044748094346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,1024,0.030461334519916113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,2048,0.02975822157329983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,768,0.023081777824295893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,2560,0.08118489053514269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,1536,0.02532711128393809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,512,0.017500445246696472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,2048,0.06684355603324042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,1024,0.021137777302000258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,1536,0.053336001104778714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,256,0.013033777475357056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,768,0.019719999697473314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,128,0.011011555790901184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,512,0.016872000363137987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,1024,0.043980442815356784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,768,0.03362400001949734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,64,0.01036711119943195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,256,0.015791111522250705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,6144,32,0.009685333404276106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,512,0.02696888811058468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,6144,128,0.016172443826993305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,256,0.023372444841596816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,128,0.022358222140206233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,64,0.022629333866967097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,6144,32,0.022615111536449854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,16384,0.39438843727111816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,51200,0.5002737575107151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,65536,0.6543528768751357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,16384,0.42168354988098145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,16384,0.1549137830734253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,12288,0.10994133022096421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,12288,0.31101867887708873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,10240,0.24067910512288412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,65536,1.504701296488444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,12288,0.3943475617302789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,10240,0.1163075500064426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,8192,0.2152266634835137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,10240,0.25682489077250165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,8192,0.07804177867041694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,7168,0.18561689058939615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,51200,1.76293150583903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,6144,0.16032800409528944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,7168,0.07068710856967501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,5120,0.1266231139500936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,6144,0.06645333104663424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,65536,2.3773271772596574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,6144,0.1754666699303521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,5120,0.0549777779314253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,4096,0.09746311108271281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,51200,1.1183937920464408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,5120,0.13119644588894314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,3584,0.08556622266769409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,4096,0.04545955525504219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,3072,0.07446044683456421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,4096,0.10746044582790798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,3584,0.03805155555407206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,2560,0.0646266672346327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,3072,0.03631466627120972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,3072,0.08251111374961005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,2048,0.04740177922778659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,2560,0.02973777718014187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,2560,0.0691724419593811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,1536,0.037808888488345675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,2048,0.026913776993751526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,8192,0.22907643847995332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,1024,0.025262221693992615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,2048,0.058287998040517174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,768,0.02068444424205356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,7168,0.17987110879686144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,1536,0.04581244455443489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,1024,0.018583110637134977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,512,0.015863110621770222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,1024,0.0347306662135654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,256,0.011561777856614856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,768,0.017224889662530687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,512,0.015728889240158927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,128,0.009650666680600908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,512,0.02335022224320306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,256,0.014706666270891825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,64,0.008549333446555668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,256,0.02025422288311852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,128,0.014476444986131458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,5120,32,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,128,0.019573332534896005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,3584,0.09230222304662068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,64,0.01958666741847992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,32,0.019295111298561096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,5120,1536,0.022063111265500385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,65536,0.5272969139946831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,51200,0.4141440126630995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,16384,0.3112728860643175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,5120,768,0.02932888931698269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,51200,0.8357306586371528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,65536,1.101043595208062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,12288,0.22077777650621203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,16384,0.1178026729159885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,12288,0.09415555662579006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,10240,0.19766667154100206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,16384,0.39300354321797687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,10240,0.07948799928029378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,12288,0.25486310323079425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,8192,0.1584764454099867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,7168,0.13890843921237522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,10240,0.2339306672414144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,8192,0.06458133459091187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,7168,0.05974666939841377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,51200,1.3198311063978407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,6144,0.12139911121792263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,8192,0.18706399864620635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,5120,0.0972728861702813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,65536,1.7528817918565538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,7168,0.14965421623653835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,6144,0.05055555701255798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,4096,0.07239911291334364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,3584,0.06333421998553805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,5120,0.04346844553947449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,6144,0.1311848825878567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,4096,0.03819288810094198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,5120,0.11465244823031956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,3584,0.032293332947625056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,2560,0.04710666669739617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,4096,0.0896951092614068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,2048,0.03927733169661628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,3584,0.07825422286987305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,3072,0.028627554575602215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,1536,0.027592000034120347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,2560,0.026354667213228013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,3072,0.06899111138449775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,1024,0.021026665965716045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,2048,0.022667555345429316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,2560,0.05862400266859266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,768,0.01661066710948944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,1536,0.018923555811246235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,2048,0.04924177792337206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,512,0.01276711126168569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,1024,0.016120000018013846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,1536,0.03947910997602675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,256,0.009287999735938178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,1024,0.029576000240114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,768,0.015422221687104968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,128,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,512,0.01383733335468504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,768,0.02498577700720893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,64,0.006923555499977536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,512,0.02018044392267863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,32,0.007090667055712805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,128,0.012831110921170978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,256,0.017512000269360013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,128,0.016833777228991192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,64,0.01684266659948561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,4096,32,0.016504888733228047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,4096,3072,0.0544088880221049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,65536,0.44574843512641055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,51200,0.35253066486782497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,51200,0.6902764638264974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,16384,0.28206578890482586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,4096,256,0.013129777378506131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,65536,1.1888595157199435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,16384,0.10745866431130303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,12288,0.2133315536710951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,16384,0.3450942304399278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,12288,0.11852888266245525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,10240,0.1688222222858005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,51200,1.1241093741522896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,12288,0.2222391234503852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,65536,1.5617297490437825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,10240,0.07157066795561048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,8192,0.18680177794562447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,7168,0.13079110781351724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,8192,0.06019733349482218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,10240,0.2554408974117703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,7168,0.051170667012532554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,6144,0.1443155606587728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,8192,0.16671466827392578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,5120,0.08996178044213189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,6144,0.045773333973354764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,4096,0.07007644573847453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,7168,0.15739822387695312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,5120,0.044512000348832875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,3584,0.052947554323408336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,6144,0.11512888802422418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,4096,0.03226844469706217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,5120,0.10013066397772895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,2560,0.04310844341913859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,3584,0.028526223368114893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,4096,0.08210488822725084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,3584,0.07211822271347046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,3072,0.026055999928050574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,1536,0.02713599966631995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,3072,0.06108711163202921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,2560,0.02408444384733836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,1024,0.019355555375417072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,2560,0.052763554784986705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,2048,0.02050755586889055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,768,0.015160888433456421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,2048,0.044417778650919594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,1536,0.018087999688254464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,512,0.011395555403497485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,1024,0.015255999233987598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,1536,0.03552800085809495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,256,0.008661333057615492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,768,0.01427111029624939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,1024,0.02646666599644555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,512,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,128,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,768,0.022337777747048274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,256,0.01239111108912362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,512,0.018057778477668762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,64,0.006457777900828256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,32,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3584,128,0.012499555945396423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,256,0.015832889411184523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,128,0.015216888652907478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,64,0.015138667490747241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,3072,0.05486311184035408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3584,2048,0.03517155514823066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,65536,0.40501422352261013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,51200,0.32769332991706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,16384,0.2119342221154107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,65536,0.8681066830952963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,51200,0.8503715727064344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,12288,0.16022222571902806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,16384,0.09762044747670491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,16384,0.3221048778957791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,10240,0.14588710996839735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,12288,0.09289777941173977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,12288,0.21804711553785536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,8192,0.12250399589538574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,51200,0.9782159593370227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,10240,0.07586933506859674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,65536,1.2332337697347004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,10240,0.1711404456032647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,7168,0.10285955667495728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,6144,0.0979057749112447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,8192,0.05206400156021118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,7168,0.04863111178080241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,5120,0.08121066623263888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,8192,0.18932799498240152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,6144,0.04195822278658549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,7168,0.12353777885437012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,5120,0.03419999943839179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,4096,0.06836889187494914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,6144,0.1039866672621833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,3584,0.05335466729270088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,5120,0.0887848867310418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,4096,0.029170665476057265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,3072,0.04614488946066963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,3584,0.025753777888086107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,4096,0.07073511017693414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,3072,0.0239964442120658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,3584,0.0628897746404012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,2048,0.0335457788573371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,2560,0.020862221717834473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,3072,0.05465066764089796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,1536,0.022481777601771887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,2048,0.018604444132910836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,1024,0.016557332542207506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,2048,0.0403111113442315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,1536,0.016372443901167977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,768,0.013156444662147097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,1536,0.03191466795073615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,1024,0.014167111780908374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,512,0.010729778144094678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3584,32,0.015559110376569959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,1024,0.02404088940885332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,768,0.01329066687160068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,256,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,768,0.020356444848908316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,512,0.012257777982287936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,128,0.007231111327807109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,512,0.016484444340070088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,256,0.011374221907721626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,64,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,256,0.014339556296666464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,32,0.006508444332414203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,3072,128,0.011647111011876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,128,0.013782222237851886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,3072,2560,0.039891554249657526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,64,0.013596444494194455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,32,0.013839999834696451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,3072,2560,0.047093331813812256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,65536,0.35859200689527726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,51200,0.28146666950649685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,16384,0.21291732788085938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,51200,0.6766355302598742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,65536,0.8819937705993652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,12288,0.16197155581580266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,16384,0.0875955555174086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,10240,0.14229955938127306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,16384,0.24494756592644584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,12288,0.08502578073077732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,51200,0.7938870853847928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,65536,0.9916221830579969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,8192,0.11519732740190293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,10240,0.059782220257653125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,12288,0.1726684437857734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,7168,0.09128799703386094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,6144,0.07575288746092054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,10240,0.15404621760050455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,5120,0.052910221947564014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,7168,0.058634665277269155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,6144,0.03717422154214647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,4096,0.04969244533114963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,7168,0.10753333568572998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,6144,0.08949510918723212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,3584,0.03858222232924567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,5120,0.03182755576239692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,3072,0.03268622358640035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,4096,0.02872177627351549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,5120,0.07942310969034831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,2560,0.030960000223583643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,4096,0.06272177563773261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,3584,0.025472000241279602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,2048,0.02551022171974182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,3072,0.023710222707854375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,3584,0.05429688758320279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,2560,0.019832889238993328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,1536,0.0196426659822464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,3072,0.04844533403714498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,1024,0.014074666632546319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,2048,0.017495110630989075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,2560,0.04177688890033298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,768,0.011320888996124268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,1536,0.015232889188660515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,2048,0.0355040000544654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,512,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,1024,0.013415999710559845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,1536,0.02881866693496704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,256,0.006351111249791251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,1024,0.02163377735349867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,512,0.011415111521879831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,768,0.01800355480776893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,128,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,512,0.014898666077189975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,256,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,64,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2560,32,0.0058408888677755995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,128,0.011008000208271874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,128,0.012239110966523489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,64,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,32,0.012395555774370829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,65536,0.6311457951863607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,65536,0.30253511004977757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,768,0.012727999852763282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2560,8192,0.06409155660205416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,65536,0.8319644398159451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,256,0.0127368892232577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,16384,0.15572710831960043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,51200,0.52019288804796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,51200,0.24176979064941406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,12288,0.11564711729685466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,16384,0.07636888821919759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,12288,0.059756444560156934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,51200,0.6611386405097114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,16384,0.19737866189744738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,8192,0.07460088862313165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,12288,0.20768354998694527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,10240,0.051867554585138954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,10240,0.1253066725201077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,7168,0.08460799853006999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,8192,0.040877332290013633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,6144,0.06011199951171875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,8192,0.10142844253116184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,7168,0.03613866700066461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,5120,0.044960889551374644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,7168,0.08870755301581489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,6144,0.036970665057500206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,4096,0.0368115570810106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,3584,0.03219822380277846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,5120,0.0277493331167433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,6144,0.07856444517771403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,3072,0.0299928883711497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,4096,0.024022221565246582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,5120,0.06499822272194757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2560,8192,0.12008978260887994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,2560,0.024902222885025874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,4096,0.052836444642808705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,2048,0.02019466625319587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,3072,0.02004888819323646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,3584,0.047741333643595375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,1536,0.017302221722073026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,2560,0.017986666825082567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,3072,0.042485333151287503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,2048,0.016174222032229107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,10240,0.09265422158771092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,2560,0.03649333450529311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,1536,0.014377777775128683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,768,0.010394666757848527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,2048,0.030971556901931763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,512,0.008440000315507254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,1536,0.024837333295080397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,1024,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,768,0.011983111500740051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,256,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,1024,0.018942221999168396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,128,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,512,0.011074666347768573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,768,0.016029333074887592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,64,0.004518222063779831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,256,0.010440888504187265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,512,0.013181333740552267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,32,0.004783111313978831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,128,0.010423999693658618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,256,0.011349333657158745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,128,0.01075466639465756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,64,0.010686222049925061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,2048,32,0.011031111081441244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,2048,3584,0.021798221601380244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,65536,0.25001955032348633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,65536,0.4930551317003038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,16384,0.11058488819334243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,51200,0.19795378049214682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,12288,0.08067555560006036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,2048,1024,0.012281777958075205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,16384,0.07730222410625882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,10240,0.08220977915657891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,16384,0.16513333055708143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,12288,0.054302222198910184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,51200,0.5179164674546983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,8192,0.06615289052327473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,12288,0.12450399663713242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,10240,0.046091556549072266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,7168,0.05486311184035408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,10240,0.11698488394419353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,8192,0.03416266706254747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,7168,0.030525333351559106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,8192,0.08420888582865398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,6144,0.02734755476315816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,7168,0.07425689034991793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,4096,0.030054221550623577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,6144,0.06605066855748494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,5120,0.024513777759340074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,3584,0.02609777781698439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,5120,0.05430755681461758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,4096,0.020956445071432326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,3072,0.022836445106400385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,4096,0.0450764430893792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,51200,0.39652445581224227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,3584,0.018606222338146634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,2560,0.018992889258596633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,3584,0.040418667925728693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,3072,0.017264889346228707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,2048,0.015896888242827523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,3072,0.03607199920548333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,65536,0.6645155482821995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,2560,0.015812445018026564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,1536,0.012531555361217923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,2560,0.031014220582114324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,1024,0.009584888815879822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,2048,0.014461333552996317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,1536,0.01311466677321328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,2048,0.026219555073314246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,6144,0.048713776800367564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,768,0.007696888513035244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,512,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,1024,0.011396444506115384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,5120,0.03603733248180813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,1024,0.016435555285877652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,768,0.010767999622556897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,256,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,512,0.010381333529949188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,128,0.004569777597983678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,768,0.013920888304710388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,512,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,64,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1536,32,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,128,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,256,0.010019555687904358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,128,0.009672889278994666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,64,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,32,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,65536,0.3091031180487739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,65536,0.19448799557156035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,51200,0.23211465941535103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,16384,0.07994755771425036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,51200,0.17438222302330864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,65536,0.5157733493381077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,12288,0.06161866585413615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,16384,0.05343022280269199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1536,1536,0.021369778447681006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,51200,0.4058826764424642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,16384,0.13310310575697157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,10240,0.05211911267704434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,8192,0.039593779378467135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,12288,0.04434222314092848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,7168,0.03566133313708835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,10240,0.03661244445376926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,12288,0.10049511326683892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1536,256,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,8192,0.027744000156720478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,6144,0.029639999071757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,10240,0.08456622229682074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,5120,0.02456355591615041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,8192,0.06797422303093804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,7168,0.025736000802781846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,4096,0.01943466729587979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,6144,0.022200889057583276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,3584,0.01759466694460975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,7168,0.059897780418395996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,5120,0.01961688862906562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,3072,0.015616888801256815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,4096,0.01722577710946401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,6144,0.052208890517552696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,2560,0.013740444348918067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,5120,0.04449955622355143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,3584,0.01588711142539978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,2048,0.0116462219092581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,4096,0.03681955403751797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,3072,0.014478221535682678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,1536,0.009356444080670675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,3584,0.03326755430963304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,2560,0.013530666629473368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,3072,0.02941688895225525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,1024,0.007531555162535773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,2048,0.012395555774370829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,2560,0.025418667329682246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,768,0.006528888725572162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,1536,0.011122666299343109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,2048,0.021885333789719477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,1024,0.010281778044170803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,1536,0.01789688898457421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,768,0.010049777726332346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,256,0.00425155543618732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,1024,0.013791999883121915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,128,0.0035537779331207275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,512,0.009330666727489894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,768,0.011843555503421359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,64,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,32,0.003535110917356279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,256,0.008966222405433655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,512,0.010238221950001186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,1024,128,0.00869955536392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,256,0.008633777499198914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,128,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,64,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,65536,0.22022577126820883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,1024,32,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,16384,0.06867911418279012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,65536,0.1930462254418267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,51200,0.15356355243259007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,16384,0.05304355422655741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,65536,0.4349208937750922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,51200,0.34630754258897567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,16384,0.1197777721616957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,10240,0.03897422220971849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,12288,0.042077332735061646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,8192,0.030419554975297716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,1024,512,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,10240,0.036328001154793635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,12288,0.08587644497553508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,7168,0.02738933265209198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,10240,0.07305777735180326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,8192,0.027056889401541814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,6144,0.023531556129455566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,7168,0.024297777149412367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,8192,0.058168888092041016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,5120,0.020032000210550096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,7168,0.051813334226608276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,6144,0.02223377757602268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,4096,0.015944000747468736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,6144,0.045273777511384755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,5120,0.019496000475353666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,3584,0.014212444424629211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,4096,0.016926222377353244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,5120,0.03841777642567953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,51200,0.2069244384765625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,3072,0.012950222525331708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,2560,0.011301333705584208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,4096,0.03201244605912103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,3584,0.015684444043371413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,2048,0.009312888814343346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,3072,0.014468444718254937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,3584,0.028639998700883653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,1536,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,12288,0.04890933301713732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,2560,0.013419555293189155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,3072,0.025348444779713947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,1024,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,2560,0.022239110536045496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,2048,0.012337778177526263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,768,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,2048,0.01905066602759891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,1536,0.011232888533009423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,512,0.004538666870858935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,1024,0.010302222437328762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,1536,0.015665777855449252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,256,0.0035626664757728577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,768,0.009834667046864828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,1024,0.012263110942310758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,128,0.0031759999692440033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,512,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,768,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,256,0.008984888593355814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,64,0.003165333428316646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,512,0.0092284447617001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,768,32,0.0032088889016045462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,768,128,0.00869066682126787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,256,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,128,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,64,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,65536,0.2205991082721286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,768,32,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,51200,0.16796444522009957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,65536,0.1469795571433173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,16384,0.057018664148118764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,12288,0.04250399933920967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,10240,0.03457333313094245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,16384,0.04153155618243747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,51200,0.11800889174143474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,12288,0.033246222469541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,8192,0.026551110876931086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,65536,0.38489688767327207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,16384,0.10635111067030166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,51200,0.3022150993347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,10240,0.029205332199732464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,7168,0.022963555322753057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,12288,0.0767493314213223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,6144,0.020229332976871066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,8192,0.0203155560625924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,10240,0.06428355640835233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,5120,0.017903111047214933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,4096,0.014471999473041959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,7168,0.01866933372285631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,6144,0.01711466742886437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,3584,0.01315822203954061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,5120,0.01514933341079288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,7168,0.04685511191685995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,6144,0.04066133168008592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,4096,0.013243555194801755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,5120,0.03427822060055203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,2560,0.010068444742096795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,4096,0.028773334291246202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,3584,0.012440889245933957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,2048,0.0087013335691558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,3072,0.011753777662913004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,3584,0.02572711143228743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,1536,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,3072,0.02286311156219906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,2560,0.011196444431940714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,2048,0.01053955571519004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,2560,0.02029244436158074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,1024,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,1536,0.009716444545321995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,2048,0.017435555656750996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,768,0.005009777843952179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,1536,0.014369777507252164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,1024,0.009338666995366415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,512,0.004187555362780889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,768,0.008818666968080733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,1024,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,256,0.003678222083383136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,768,0.010090666512648264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,512,0.008368888662921058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,128,0.0031884445084465873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,256,0.00793688909875022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,512,0.008839999636014303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,64,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,32,0.0028942223224374983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,512,128,0.008018666671382057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,8192,0.05178755521774292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,128,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,512,3072,0.011448000040319232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,64,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,65536,0.12675733036465114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,32,0.00638488887084855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,51200,0.10209777620103623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,16384,0.03867288761668735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,65536,0.14485777748955622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,51200,0.11678844028049046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,12288,0.026214222113291424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,65536,0.3116684489780002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,16384,0.040821333726247154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,10240,0.020025778147909377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,51200,0.2439626587761773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,12288,0.03218400147226121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,16384,0.08197866545783149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,8192,0.016326222154829238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,7168,0.014717333846622042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,10240,0.028697778781255085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,12288,0.06197244591183133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,6144,0.013103110922707452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,10240,0.05197244551446703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,7168,0.018240888913472492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,8192,0.042295111550225147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,5120,0.011563555234008364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,7168,0.03790666659673055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,6144,0.016728889611032274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,4096,0.009670221971140968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,6144,0.033398220936457314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,5120,0.014987554815080432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,3584,0.008595555192894405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,4096,0.013110222087966071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,5120,0.028202666176689997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,3072,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,4096,0.023414222730530634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,3584,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,2560,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,3584,0.021317332983016968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,3072,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,2048,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,1536,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,2560,0.011108444796668159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,3072,0.01898755629857381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,1024,0.00423911131090588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,2560,0.016910221841600206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,1536,0.009721777505344814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,768,0.0038844446341196695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,2048,0.014451555079884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,1536,0.012156444291273752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,512,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,1024,0.010006222460005019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,768,0.00871377769443724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,256,0.0028479999552170434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,8192,0.019778667224778067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,768,0.008719111482302347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,512,0.00831200016869439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,128,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,512,0.0075804442167282104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,64,0.002567111204067866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,256,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,512,256,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,128,0.00562755556570159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,65536,0.08248622549904717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,64,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,256,32,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,51200,0.06654933426115248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,65536,0.1451679997973972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,65536,0.2676604323916965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,1024,0.009336888790130615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,51200,0.11611821916368271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,51200,0.21062400605943468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,12288,0.020875554945733812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,16384,0.041505777173572116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,16384,0.07054489188724093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,256,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,10240,0.016896888613700867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,128,0.007825777762466008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,12288,0.03353688783115811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,256,32,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,12288,0.05382844474580553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,8192,0.012858666479587555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,7168,0.011471110913488599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,10240,0.04518489042917887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,8192,0.021290666527218286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,6144,0.010207999911573198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,8192,0.03698933455679152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,7168,0.019010666343900893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,5120,0.0086666668454806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,7168,0.03277511066860623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,6144,0.01755288905567593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,256,2048,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,4096,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,6144,0.028786665863460962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,5120,0.015600000818570455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,3584,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,16384,0.027051554785834417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,5120,0.024824889169798955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,4096,0.013818666338920593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,3584,0.013063111239009433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,4096,0.020848888489935134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,2560,0.00552977787123786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,3584,0.018957333432303537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,2048,0.004562666846646203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,3072,0.016900445024172466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,2560,0.0151555554734336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,1536,0.0041946665280395085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,2048,0.013155555559529198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,1536,0.009972444838947719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,1536,0.011024889018800525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,1024,0.00350577798154619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,1024,0.009047110875447592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,768,0.0031662223239739737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,1024,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,3072,0.005965333431959152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,768,0.008970666262838576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,512,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,768,0.008212444682916006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,256,0.002505777817633417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,3072,0.012153777811262341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,512,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,128,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,2560,0.011366222467687396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,64,0.0021013333979580137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,512,0.0070462218589252895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,256,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,128,32,0.0021226666867733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,256,0.005780444376998478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,2048,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,128,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,65536,0.0789653327729967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,10240,0.02881422307756212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,51200,0.06321777900060017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,16384,0.022973333795865376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,64,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,12288,0.014169777433077494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,128,32,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,10240,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,8192,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,7168,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,16384,0.07042133145862155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,65536,0.26807644632127553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,12288,0.05340533455212911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,6144,0.009372444616423713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,51200,0.21039111084408232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,10240,0.04510400030348036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,5120,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,8192,0.03667111198107401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,4096,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,3584,0.0064204442832205035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,3072,0.0058453331391016645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,2560,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,7168,0.03261422117551168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,2048,0.004550222307443619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,6144,0.02868622210290697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,5120,0.02467022173934513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,1536,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,4096,0.0206995556751887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,1024,0.0032257777121331957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,3584,0.01893422173129188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,3072,0.016880000631014507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,768,0.002879111096262932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,512,0.0026008888251251644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,2560,0.015117333994971381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,256,0.0024702221983008915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,128,0.0021217777911159727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,2048,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,64,0.0021840000732077495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,1536,0.011114666859308878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,1024,0.009020444419648912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,64,32,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3585,128,128,0.00794400026400884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,512,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,256,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,65536,0.07589955462349786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,51200,0.05928444200091892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,128,0.005304000029961268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,64,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,12288,0.012553777959611682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,16384,0.02068088948726654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,10240,0.010964444114102257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,32,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,8192,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,16384,0.07022400034798516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,65536,0.2680995464324951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,12288,0.053661333190070264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,7168,0.008983111215962304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,10240,0.045250667466057666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,6144,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,5120,0.010647999743620554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,8192,0.03669510947333442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,4096,0.008808888494968414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,3584,0.008088889221350351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,7168,0.03258488906754388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,6144,0.028564443190892536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,3072,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,5120,0.024717332588301763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,2560,0.006344888773229387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,2048,0.004523555437723796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,4096,0.020729778541458976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,3584,0.01901955571439531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,3072,0.017215111189418368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,2560,0.015433778365453085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,64,768,0.008070222205585903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,1024,0.0032106666929192017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,2048,0.013140444126394061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,768,0.0028773333049482773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,512,0.00254488881263468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,1536,0.011153777440388998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,256,0.0022044444663657085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,128,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,1024,0.00904444439543618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,768,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,512,0.007150222029950883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,32,0.0018871112002266778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,256,0.005640888793600931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,128,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,51200,0.21029867066277397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,64,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3585,32,32,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,1536,0.003875555677546395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3585,32,64,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,16384,1.4759332868787978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,16384,4.362159305148654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,16384,5.318661159939237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,12288,3.2965901692708335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,51200,5.5997467041015625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,16384,4.481923421223958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,12288,1.0790480507744682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,12288,3.60715823703342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,10240,0.904854244656033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,10240,2.8936356438530813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,10240,3.506842719184028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,12288,3.265216827392578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,8192,0.7446435292561849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,10240,2.648103925916884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,8192,2.3649590810139975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,8192,2.236490673489041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,7168,0.639178646935357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,8192,2.160206264919705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,51200,13.775572035047743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,7168,2.1967457665337458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,7168,1.9423696729871962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,6144,0.5874648623996311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,7168,1.9330480363633897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,6144,1.7482692930433485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,6144,1.6983493169148762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,5120,1.4058169258965387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,5120,0.49445777469211155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,6144,1.537941296895345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,5120,1.428884400261773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,4096,1.248311996459961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,4096,0.3864560127258301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,4096,1.1293492847018771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,5120,1.3808871375189886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,3584,0.3428835603925917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,3584,1.0510924657185872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,4096,1.0816693835788302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,3072,0.840460459391276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,3584,1.021807140774197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,3584,0.9050683975219727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,3072,0.3059111171298557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,3072,0.8413297865125867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,2560,0.7531173494127062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,2560,0.26827910211351186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,2560,0.713680002424452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,3072,0.7903173234727648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,2048,0.5392586919996474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,2048,0.22825866275363496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,2560,0.6729413138495551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,1536,0.4320426781972249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,2048,0.601070245107015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,1536,0.1941386726167467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,2048,0.5476711061265733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,1024,0.3146097660064697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,1536,0.49149688084920246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,1536,0.4260719882117377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,768,0.24161601066589355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,1024,0.14847377936045328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,1024,0.34796799553765195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,1024,0.3258826732635498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,768,0.27480355898539227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,51200,15.267144097222221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,768,0.2712471220228407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,256,0.12030577659606934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,512,0.23470576604207358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,512,0.1465173297458225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,512,0.20958311027950713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,128,0.09596000115076701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,256,0.13434933291541204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,256,0.20939111709594727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,256,0.16557333204481336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,64,0.08570222059885661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,128,0.19723999500274658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,32,0.08355466524759929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,768,0.12757689423031277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,51200,21.98474629720052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,64,0.20210311147901747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,65536,32,0.2030773295296563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,65536,128,0.09723733531104194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,65536,512,0.16583466529846191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,65536,128,0.12503200107150608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,51200,4.420947604709201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,65536,5.743069542778863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,16384,4.200839996337891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,51200,10.853506300184462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,16384,4.2567261589898004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,16384,1.1355315314398873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,65536,13.661756727430555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,12288,2.9641439649793835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,12288,0.8872417873806424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,16384,3.6203464931911893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,12288,2.945309321085612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,51200,11.945353190104166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,12288,2.5254942576090493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,51200,17.235788981119793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,10240,0.7070382436116537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,10240,2.7285448710123696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,10240,2.301983939276801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,8192,1.9299973381890192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,8192,0.6782631344265408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,10240,2.168633778889974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,65536,14.6895751953125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,8192,1.8509769439697266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,7168,1.6510640250311956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,7168,0.49329068925645614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,8192,1.6671981811523438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,7168,1.5263857311672633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,6144,0.4654284583197699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,6144,1.5449395709567602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,6144,1.3130017386542427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,7168,1.453078269958496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,65536,21.820929633246525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,5120,1.0723910861545138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,5120,0.41852445072597927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,6144,1.2749431398179796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,5120,1.0756461885240343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,4096,0.30738666322496205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,4096,0.9293555153740777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,5120,1.0390675862630208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,3584,0.2758222156100803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,3584,0.8063617812262641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,4096,0.7821271154615613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,3584,0.8062017758687338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,3072,0.7209084298875598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,3584,0.7025902006361219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,3072,0.2741484377119276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,3072,0.7044969134860568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,2560,0.536644458770752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,2560,0.21159733666314018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,3072,0.6191031138102213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,2048,0.45006222195095485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,2560,0.5640488730536567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,2048,0.18380978372361925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,2560,0.5238844553629557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,4096,0.8836648729112414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,1536,0.34094667434692383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,2048,0.4649102422926161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,1536,0.1462364461686876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,1024,0.22385954856872559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,2048,0.44226932525634766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,768,0.17971289157867432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,1024,0.11691910690731472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,1536,0.5311342345343696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,1024,0.2708115577697754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,1536,0.3434266779157851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,512,0.1499626636505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,768,0.23159644338819715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,1024,0.3044568962521023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,768,0.10247555706236099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,512,0.08949866559770371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,768,0.21108799510531953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,512,0.22928176985846627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,128,0.07618044482337104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,512,0.16757599512736002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,256,0.07951822545793322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,64,0.06386488676071167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,256,0.16666221618652344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,256,0.13532622655232748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,51200,128,0.07743466562694974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,32,0.06581955485873751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,128,0.15862488746643066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,51200,128,0.10328266355726455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,64,0.15867200162675646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,51200,32,0.16092799769507513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,51200,256,0.0963111122449239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,65536,1.7847910987006292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,16384,1.2830479939778645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,51200,3.5466560787624783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,51200,1.308834711710612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,65536,4.790679931640625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,16384,0.38491378890143496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,16384,1.2804844114515517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,16384,1.176161766052246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,12288,0.9591822094387479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,51200,3.838997311062283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,12288,0.3016773329840766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,12288,0.9463866551717123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,12288,0.7992400063408746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,10240,0.8396115303039551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,10240,0.23520000775655112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,65536,6.869778527153863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,65536,5.198193868001302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,8192,0.6335528691609701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,10240,0.779544883304172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,8192,0.18733955754174125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,10240,0.6456089019775391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,7168,0.165248884095086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,8192,0.5540684594048394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,8192,0.5914275381300185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,6144,0.43320266405741376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,7168,0.5493137571546767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,7168,0.5333013534545898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,7168,0.4549119737413194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,6144,0.1530133353339301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,51200,6.413233015272353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,5120,0.36998576588100857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,5120,0.1278053389655219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,6144,0.43181512090894913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,4096,0.3049768871731228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,6144,0.39757511350843644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,4096,0.10690310928556655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,3584,0.2407848834991455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,4096,0.29028532240125865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,5120,0.48875824610392254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,5120,0.3303137885199653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,3584,0.09604799747467041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,3072,0.21344355742136636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,3584,0.27066487736172146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,3072,0.08415999677446152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,2560,0.18016533056894937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,3584,0.25587556097242564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,3072,0.22057510746849907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,2560,0.07353955507278442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,3072,0.20935377809736463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,2048,0.14794222513834634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,2560,0.1965928872426351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,1536,0.11983999941084121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,2048,0.06281510988871257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,2560,0.1803706619474623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,2048,0.15555466545952693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,1024,0.08038133382797241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,2048,0.14845066600375706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,1536,0.05220444334877861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,1536,0.1261377731959025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,768,0.06849600209130181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,1024,0.04409155580732557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,1536,0.12080800533294678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,1024,0.09480355845557319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,512,0.053379555543263756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,768,0.03973244296179877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,768,0.07481600178612603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,1024,0.09240977631674872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,256,0.03800444470511542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,768,0.07877866427103679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,512,0.03599111239115397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,512,0.06366399923960368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,256,0.03214755654335022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,128,0.02718222141265869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,4096,0.27877243359883624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,512,0.06384799877802531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,64,0.02259822189807892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,256,0.05719821982913547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,16384,32,0.02390844457679325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,16384,128,0.03171644608179728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,256,0.0506551100148095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,128,0.05434044533305698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,16384,128,0.03841244512134128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,32,0.05601333247290718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,16384,64,0.05485422081417508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,65536,1.22743648952908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,16384,1.0797582202487521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,51200,0.9765467113918729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,51200,2.791049745347765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,65536,3.3688151041666665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,16384,0.2798239919874403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,12288,0.6597929000854492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,16384,0.9872266981336805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,51200,2.9594470130072703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,51200,4.17287360297309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,12288,0.6970213254292806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,16384,0.8663093249003092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,12288,0.23291643460591635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,10240,0.5539369053310818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,10240,0.1810213327407837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,65536,5.106384701199002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,65536,3.8173120286729603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,10240,0.5640337732103136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,12288,0.6127582126193577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,10240,0.5047386487325033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,8192,0.4651466475592719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,8192,0.14935021930270725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,7168,0.37795111868116593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,8192,0.4656808641221788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,7168,0.12786932786305746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,7168,0.39760711457994247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,8192,0.41513599289788145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,6144,0.34081066979302305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,7168,0.3569511042700873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,6144,0.1123431126276652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,5120,0.2714017762078179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,6144,0.3646204471588135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,5120,0.09886222415500218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,4096,0.19777866204579672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,6144,0.30606044663323295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,4096,0.08263822396596272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,5120,0.2698631021711561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,4096,0.23977687623765734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,4096,0.21201777458190918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,3584,0.07440355751249525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,3072,0.1610062254799737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,3584,0.19426311386956108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,3072,0.06589066982269287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,3072,0.17389599482218424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,3584,0.1860213279724121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,3072,0.16383822758992514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,2048,0.09752266936832005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,2560,0.1449191172917684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,2560,0.05774933099746704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,1536,0.07546399699317084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,2560,0.14014400376213923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,2048,0.04968000120586819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,2048,0.1193386713663737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,5120,0.2976693312327067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,2048,0.11891555786132812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,1536,0.041675554381476514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,1024,0.06002044677734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,3584,0.1685697767469618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,1536,0.09488622347513835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,768,0.052573333183924355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,1024,0.03554666704601712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,1536,0.09548088577058579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,768,0.0322053333123525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,1024,0.07243555784225464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,1024,0.07458933194478352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,768,0.05834488736258613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,768,0.06260089079538982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,256,0.03308088911904229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,512,0.029525332980685767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,512,0.04980977707438999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,2560,0.12365244494544135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,128,0.02829422222243415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,256,0.026660444007979497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,512,0.05174755718972948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,256,0.04434222314092848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,64,0.025713778204388086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,12288,128,0.02608799934387207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,32,0.026760000321600173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,128,0.04231022132767571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,128,0.031174222628275555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,64,0.04273955689536201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,12288,32,0.04313866628540886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,12288,512,0.044300443596310086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,12288,256,0.040793776512145996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,65536,1.1076551013522677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,16384,0.8394604259067111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,51200,0.8704506556193033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,51200,2.4737741682264542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,65536,2.828048917982313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,16384,0.26324444346957737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,12288,0.5784942309061686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,16384,0.8693964216444227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,51200,3.40791384379069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,16384,0.7181537946065267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,12288,0.614699575636122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,12288,0.1849128935072157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,51200,2.492295159233941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,10240,0.45420000288221574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,65536,4.296151055230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,65536,3.2094391716851125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,12288,0.5214942296346029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,10240,0.51364533106486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,8192,0.39519643783569336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,8192,0.14104444450802273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,10240,0.4281911055246989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,7168,0.3358328872256809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,8192,0.4392080042097304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,7168,0.12819466325971815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,8192,0.3466302288903131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,7168,0.3290639983283149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,6144,0.09695021973715888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,7168,0.4016115665435791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,5120,0.21402488814459908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,6144,0.41944087876213926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,6144,0.26365778181287974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,5120,0.08386577500237359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,10240,0.15643821822272405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,4096,0.22767912016974556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,5120,0.23899022738138834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,4096,0.07059733072916667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,3584,0.19961955812242296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,4096,0.18606132931179473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,5120,0.24030311902364096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,4096,0.18151556120978463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,3072,0.11500710911220974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,3584,0.06577333476808336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,3584,0.16471732987297907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,2560,0.12228711446126302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,6144,0.2830977704789904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,3584,0.16079910596211752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,3072,0.14384355809953478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,2048,0.09348977936638726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,3072,0.14195911089579263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,2560,0.05009688933690389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,2560,0.12355289194318984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,1536,0.0686017804675632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,2048,0.10148621930016412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,2048,0.10250844558080037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,1536,0.0365573333369361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,1024,0.04875822199715508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,1536,0.08137510882483588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,1536,0.08379733562469482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,768,0.03646844294336107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,1024,0.030646221505271062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,1024,0.061962664127349854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,1024,0.0646373364660475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,512,0.026346666945351496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,768,0.02825688984658983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,768,0.05013422171274821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,3072,0.055705779128604464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,768,0.05486311184035408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,512,0.025528889563348558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,512,0.04195644458134969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,128,0.018917333748605516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,512,0.045653333266576133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,256,0.022895110978020564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,256,0.03738311264250014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,64,0.01587288909488254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,2560,0.12614666091071233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,128,0.02187999917401208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,32,0.01590488851070404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,128,0.03530044356981913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,128,0.027474666635195415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,64,0.03547022077772353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,10240,32,0.03535733289188809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,10240,256,0.019893333315849304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,10240,256,0.03596533338228861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,10240,2048,0.04337066743108961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,65536,0.8441475762261285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,51200,0.6742977566189237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,51200,1.9555670420328777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,65536,2.5666238996717667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,16384,0.2001031107372708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,16384,0.6720870865715874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,16384,0.5947537951999241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,12288,0.4900035328335232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,16384,0.8066337903340658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,65536,3.5166212717692056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,51200,2.8915458255343967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,65536,2.668606228298611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,12288,0.15258044666714138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,12288,0.4865848753187392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,51200,2.075639088948568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,10240,0.386919101079305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,10240,0.1424071126514011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,12288,0.43109599749247235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,8192,0.31808800167507595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,10240,0.4133031103346083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,8192,0.11480266518063015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,7168,0.26337599754333496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,10240,0.36187111006842715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,8192,0.3069262239668104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,7168,0.09323733382754856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,6144,0.22093689441680908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,7168,0.2796062098609077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,8192,0.315192010667589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,6144,0.0809440016746521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,5120,0.1877128945456611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,4096,0.14941244655185276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,6144,0.25261155764261883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,5120,0.07054933574464586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,7168,0.3359164396921794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,6144,0.22178933355543348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,3584,0.1144968933529324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,5120,0.21379289362165663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,4096,0.059458666377597384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,5120,0.18916711542341444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,3072,0.0971377756860521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,4096,0.15455378426445857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,3584,0.05364888906478882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,3584,0.15074311362372503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,4096,0.1714480055703057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,2560,0.08992977937062581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,3072,0.04756977823045519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,3584,0.13683644930521646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,3072,0.11894310845269097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,2048,0.0659102201461792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,2560,0.04261777798334757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,1536,0.0558533337381151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,2560,0.1032186680369907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,2048,0.03614044520590041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,2560,0.11572444438934326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,2048,0.08403999937905206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,1024,0.039763556586371526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,1536,0.029707554313871596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,2048,0.08814044131173028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,768,0.033639109796947904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,1536,0.06682222419314914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,1024,0.024493333366182115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,1536,0.07113066646787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,1024,0.05073777834574381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,512,0.027287999788920086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,1024,0.055832001898023814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,768,0.02253244486120012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,256,0.020046222541067336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,768,0.04128177629576789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,768,0.047001779079437256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,512,0.020801777640978496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,512,0.033379554748535156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,128,0.014739554789331226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,512,0.03943644298447503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,256,0.01847111185391744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,256,0.02902311086654663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,64,0.012392889294359418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,256,0.03139200144343906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,8192,32,0.014458666245142618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,8192,128,0.018796443939208984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,128,0.02809777855873108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,3072,0.1211128897137112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,8192,128,0.02338933282428318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,64,0.028192000256644353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,8192,32,0.02825511164135403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,65536,0.8966124852498373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,51200,1.3945769204033747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,16384,0.690332465701633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,51200,0.6718871328565809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,65536,2.247392018636068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,16384,0.6059146457248265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,16384,0.225853337181939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,12288,0.4334079954359267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,51200,2.5040772755940757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,65536,3.0110516018337674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,16384,0.5427511003282335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,12288,0.15821599960327148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,12288,0.40891112221611875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,10240,0.35598844952053493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,12288,0.39664355913798016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,10240,0.13612000147501627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,8192,0.23671110471089682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,10240,0.36612534523010254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,8192,0.09760977824529012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,7168,0.20578932762145996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,10240,0.31699734263949925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,8192,0.3365022341410319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,8192,0.2594355477227105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,7168,0.08379111025068496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,6144,0.19187643792894152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,51200,1.7961190541585286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,7168,0.2391280068291558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,5120,0.15276977750990126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,6144,0.07389422257741292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,7168,0.2250035603841146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,6144,0.20829510688781738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,4096,0.12283555666605632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,5120,0.06623111168543498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,6144,0.1950977775785658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,65536,2.357420391506619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,5120,0.16961333486768934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,3584,0.10764622026019627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,4096,0.053492443429099194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,4096,0.13717510965135363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,5120,0.16490044858720568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,3072,0.09614044427871704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,4096,0.13920533657073975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,3584,0.047464887301127114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,2560,0.07704177829954359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,3584,0.12050844563378228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,3072,0.042055110136667885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,2048,0.05966222286224365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,3584,0.12419111198849148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,3072,0.1063599983851115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,2560,0.03787111242612203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,3072,0.10693155394660102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,2048,0.03233066532346938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,2560,0.09183644586139256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,2048,0.0748951103952196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,1024,0.0362408889664544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,2048,0.07951733138826159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,1536,0.060334225495656334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,1536,0.027564444475703772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,768,0.026733333865801494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,1536,0.06410044431686401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,1024,0.022338666849666174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,1024,0.0453688899676005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,512,0.018933332628673978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,1024,0.05003822180959913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,768,0.020101333657900494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,768,0.03774044579929776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,256,0.018107554978794523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,768,0.04277688927120633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,512,0.01849244369400872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,512,0.0299635562631819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,128,0.01183111137813992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,512,0.03574488891495599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,256,0.016872000363137987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,256,0.026103110777007207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,64,0.013733333183659447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,256,0.027592000034120347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,7168,128,0.016548444827397663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,128,0.025048888391918604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,32,0.014455111490355598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,7168,128,0.02141511109140184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,64,0.025030222203996446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,32,0.02519911030928294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,65536,0.7361529138353137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,7168,1536,0.04917422268125746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,51200,1.422825813293457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,65536,1.9417093065049913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,51200,0.5689662297566732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,16384,0.47675466537475586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,7168,2560,0.09116533067491318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,16384,0.1605102221171061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,16384,0.49619197845458984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,51200,2.1806774139404297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,65536,3.1122923956977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,65536,2.08277342054579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,16384,0.5153137842814127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,12288,0.39519288804796004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,12288,0.1195031139585707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,12288,0.3573653433057997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,10240,0.29203822877671987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,10240,0.10393510924445258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,8192,0.23409777217441133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,12288,0.34719112184312606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,10240,0.3404702345530192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,8192,0.08414933416578506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,51200,1.606968879699707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,7168,0.2026302284664578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,10240,0.3181217776404487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,8192,0.2359128793080648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,6144,0.17977421813540986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,8192,0.24867555830213758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,6144,0.06492088900672065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,5120,0.1567386653688219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,6144,0.19893599881066215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,7168,0.20871733294592965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,6144,0.1848346657223172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,4096,0.10495644145541722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,5120,0.06926133235295613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,5120,0.1499306625790066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,3584,0.11063644621107314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,4096,0.04937244455019633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,4096,0.12092799610561794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,4096,0.12567555904388428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,3072,0.08535110950469971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,3584,0.0433555543422699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,3584,0.10631377829445733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,2560,0.06824177503585815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,3584,0.1211671166949802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,3072,0.038930667771233454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,3072,0.09320888916651408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,2048,0.05823733409245809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,7168,0.07482044564353095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,3072,0.0987208882967631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,7168,0.20630044407314727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,2560,0.07999022139443292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,1536,0.04533066683345371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,2560,0.08599999878141616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,2048,0.028304000695546467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,1024,0.02870755394299825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,2048,0.06618933545218574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,2048,0.07181421915690105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,1536,0.024497777223587036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,1536,0.05264977614084879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,768,0.022491556074884202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,1536,0.058896885977851025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,1024,0.020498666498396132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,1024,0.04002133342954848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,512,0.017758222089873422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,5120,0.15181599722968206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,1024,0.046021332343419395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,768,0.03286666671435038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,256,0.018041777941915724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,768,0.039577778842714094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,512,0.016928889685206942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,128,0.015384889311260648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,512,0.026483555634816486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,512,0.0325857765144772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,256,0.015460444821251763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,256,0.02276888820860121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,64,0.013785777820481194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,256,0.025507556067572698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,128,0.015857777661747403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,2560,0.03377866744995117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,6144,128,0.019609777463806998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,64,0.022309333086013794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,32,0.021951110826598272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,6144,768,0.0188746667570538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,65536,0.6015351083543565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,6144,32,0.014519999424616495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,65536,1.4568746354844835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,51200,1.0547617806328666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,51200,0.45734135309855145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,65536,2.090010748969184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,6144,128,0.022016000416543748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,65536,1.8778764936659071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,16384,0.4474293390909831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,12288,0.22792087660895455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,16384,0.13837067286173502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,16384,0.40143820974561906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,51200,1.294730716281467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,51200,1.7758009168836806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,12288,0.10319999853769939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,10240,0.24649333953857422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,12288,0.3028159936269124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,10240,0.09421955876880223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,8192,0.1854133341047499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,12288,0.29331911934746635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,10240,0.24971911642286512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,16384,0.38931822776794434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,7168,0.14501244492001003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,8192,0.07244088914659289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,10240,0.26025154855516225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,8192,0.270760006374783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,6144,0.1237564484278361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,7168,0.06493866443634033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,8192,0.198089771800571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,5120,0.10691022210650974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,7168,0.1797840065426297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,6144,0.058528886901007764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,6144,0.15650399525960287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,7168,0.17428978284200033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,5120,0.05259822143448723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,6144,0.15823733806610107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,5120,0.12750577926635742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,3584,0.07662400272157457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,4096,0.03962399893336826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,4096,0.10447911421457927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,3072,0.07116444243325128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,3584,0.036489778094821505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,4096,0.11004177729288737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,3584,0.09174844291475083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,2560,0.0574186642964681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,3584,0.09650133053461711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,3072,0.08010844389597575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,3072,0.032483554548687406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,2048,0.04495199852519565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,3072,0.08452088965309991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,2560,0.028604444530275133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,2560,0.06899733013576932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,2560,0.07544355260001288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,2048,0.024833778540293377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,2048,0.05760800176196628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,1024,0.023959999283154804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,2048,0.06263022290335761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,1536,0.021399110555648804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,1536,0.04621866676542494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,768,0.01927288870016734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,4096,0.08673333459430271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,1024,0.018231999542978074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,1024,0.034739553928375244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,512,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,5120,0.13106044133504233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,1024,0.04053777787420485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,768,0.016805332567956712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,768,0.029455112086402044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,256,0.0107342220014996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,768,0.03406933281156752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,512,0.01516355574131012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,128,0.01035822182893753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,512,0.022954665952258643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,256,0.014122666584120857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,512,0.02804622385236952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,256,0.02000711030430264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,64,0.00871733327706655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,32,0.008956444760163626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,256,0.02216000027126736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,5120,128,0.014177777700954013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,128,0.019271110494931538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,5120,1536,0.036928001377317644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,128,0.01773422294192844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,64,0.019346666004922655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,5120,32,0.019256000717480976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,5120,1536,0.05230577786763509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,65536,0.5068702167934841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,51200,0.9342169231838651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,51200,0.39234312375386554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,65536,1.2305688858032227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,16384,0.41180533832973903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,16384,0.13784444332122803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,51200,1.3063555823432076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,65536,1.6986968782212999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,51200,1.1606249279446073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,16384,0.38788265652126735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,12288,0.27533777554829914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,65536,1.5091884401109483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,12288,0.09140977594587539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,10240,0.19986932807498506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,12288,0.25520088937547475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,8192,0.15804355674319798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,10240,0.07884533537758721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,12288,0.2752435472276476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,10240,0.21243022547827828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,7168,0.14217155509524876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,10240,0.23616356319851348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,8192,0.06211022535959879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,8192,0.16817688941955566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,6144,0.12219911151462132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,7168,0.05610311031341553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,7168,0.14772355556488037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,8192,0.22312800089518228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,16384,0.34659467803107363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,6144,0.12787377834320068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,6144,0.04945688777499729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,7168,0.16623111565907797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,4096,0.08036088943481445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,5120,0.041386667225095965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,6144,0.13815200328826904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,5120,0.11228444841172959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,3584,0.06548088788986206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,5120,0.11804355515374078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,4096,0.08654044734107123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,4096,0.0347688893477122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,3072,0.05643555853101942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,4096,0.09654221932093303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,3584,0.03138755427466498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,3584,0.07749955521689521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,2560,0.049970666567484535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,3584,0.08608888917499119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,3072,0.027982221709357366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,3072,0.06758399804433186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,2048,0.037639998727374606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,3072,0.0771920018725925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,2560,0.05894755654864841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,1536,0.029543111721674602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,2048,0.021892444954978094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,2560,0.06713689035839505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,2048,0.04853511187765333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,1024,0.0211191111140781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,5120,0.0993466642167833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,2048,0.05637422535154554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,1536,0.01868799991077847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,768,0.017137777474191453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,1536,0.03877244393030802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,1024,0.015850666496488783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,512,0.013168888787428537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,1024,0.029517332712809246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,1024,0.036383999718560114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,768,0.015096000499195524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,256,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,768,0.02476177778508928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,768,0.03003466791576809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,512,0.013716444373130798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,512,0.019903111788961623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,128,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,512,0.025507556067572698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,256,0.012879999975363413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,256,0.017204445269372728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,64,0.007253333098358578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,4096,32,0.00740799970097012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,256,0.020308444897333782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,128,0.012877333495352002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,128,0.016490666402710807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,128,0.016583111551072862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,4096,2560,0.025183111429214478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,64,0.016525333126386006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,4096,32,0.016523554921150208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,4096,1536,0.046241776810752026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,65536,0.43596622678968644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,51200,0.7483341958787707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,16384,0.2896408769819471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,65536,1.2419599956936307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,51200,0.3470657666524251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,16384,0.13233066929711237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,16384,0.3564426634046767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,51200,1.1671635309855144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,65536,1.4841911527845595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,51200,1.0543058183458116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,16384,0.3279120127360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,12288,0.08113866382175021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,65536,1.3650746875339086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,10240,0.1725920041402181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,12288,0.30813688702053493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,8192,0.13741421699523926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,10240,0.06967377662658691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,12288,0.23808799849616158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,10240,0.20369244946373832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,7168,0.1264231072531806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,8192,0.06449955701828003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,8192,0.154785778787401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,10240,0.20168977313571504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,6144,0.10267288817299737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,8192,0.16457955042521158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,7168,0.0587315559387207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,7168,0.13372445106506348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,7168,0.15248621834648982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,5120,0.08630755874845718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,6144,0.04380977816051907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,6144,0.11579110887315537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,12288,0.21325155099232992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,4096,0.06624088684717815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,5120,0.03840000099605984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,5120,0.09699822134441799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,3584,0.059976001580556236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,4096,0.0314684444003635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,5120,0.10664177603191799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,3072,0.04783911175198025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,4096,0.088191999329461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,3584,0.02861600120862325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,3584,0.06947288910547893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,2560,0.04279111160172356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,3584,0.08011288775338067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,3072,0.02587644424703386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,3072,0.060982220702701144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,3072,0.07029600275887384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,2048,0.03492710987726847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,2560,0.05288711190223694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,2560,0.06053688791063097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,1536,0.026025777061780293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,2048,0.02010222276051839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,2048,0.0444124440352122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,2048,0.05161955621507433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,1536,0.017831110292010836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,1536,0.035495112339655556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,6144,0.12599910630120173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,768,0.014873777826627096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,1536,0.04239644606908163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,1024,0.02660177813635932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,4096,0.07818755838606092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,1024,0.015413333972295126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,512,0.01165688865714603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,1024,0.03250044584274292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,768,0.014133334159851074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,768,0.022944889134830896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,256,0.00978311151266098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,768,0.028015110227796767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,512,0.013072888884279462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,128,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,512,0.023372444841596816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,256,0.01239911135700014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,2560,0.023177777727444965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,256,0.015817777978049386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,64,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,256,0.019534221953815885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,32,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3584,128,0.012418666647540199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,128,0.015103111664454142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3584,128,0.01611822181277805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,64,0.0151582227812873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3584,1024,0.01945688823858897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,32,0.015145777000321282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,65536,0.3886151048872206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,51200,0.6757253540886773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3584,512,0.018295110927687753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,65536,1.0065724054972331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,51200,0.30703555213080513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,16384,0.24792265892028809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,51200,0.9745733473036023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,65536,1.2170693079630535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,12288,0.19506311416625977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,16384,0.25945599873860675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,16384,0.09244622124565972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,51200,0.9008418189154731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,65536,1.184861289130317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,10240,0.16481155819363064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,16384,0.32198044988844127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,12288,0.07919466495513916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,8192,0.13365777333577475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,10240,0.06891822152667575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,10240,0.1680426730049981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,12288,0.21618578169080946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,10240,0.19262754917144775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,7168,0.12371199660831028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,8192,0.05199911197026571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,8192,0.13804622491200766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,6144,0.10556888580322266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,8192,0.16622132725185818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,7168,0.04518577787611219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,7168,0.11769599384731716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,5120,0.07853511306974623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,6144,0.03904977771970961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,6144,0.10297244124942356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,7168,0.1315386692682902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,4096,0.06264177958170573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,6144,0.11305422253078884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,5120,0.03357244531313578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,3584,0.0547235541873508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,4096,0.028509332074059382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,5120,0.09596177604463364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,4096,0.06976177957322863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,3072,0.0455297761493259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,4096,0.07892088757620917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,3584,0.062333332167731396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,3584,0.025407110651334126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,3584,0.07136888636483087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,2560,0.040123555395338274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,3072,0.023216888308525085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,12288,0.20712355772654215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,3072,0.0543759995036655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,3072,0.06291555696063571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,2560,0.020627554919984605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,2560,0.04671644502215915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,2560,0.05531911055246989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,1536,0.026682666606373254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,2048,0.018160000443458557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,2048,0.03959644503063626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,1024,0.01754044493039449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,1536,0.016189333465364244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,2048,0.046462221278084644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,1536,0.03179555469089084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,768,0.01347644461525811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,1536,0.038872000243928694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,1024,0.013904889424641928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,5120,0.08789510859383477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,1024,0.023952000670962866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,768,0.013364444176355997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,768,0.02013511127895779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,768,0.025400888588693406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,256,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,512,0.01642933322323693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,512,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,512,0.02148977749877506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,2048,0.033302221033308245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,256,0.011411555939250521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,256,0.014163555370436775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,256,0.01790577835506863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,3072,128,0.011373333632946014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,32,0.00600000015563435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,128,0.013784888717863293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,128,0.01460355520248413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,64,0.013417777915795645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,3072,32,0.01404800017674764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,512,0.010713777608341642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,3072,1024,0.029486222399605647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,65536,0.630765332116021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,65536,0.3202906714545356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,128,0.00656355544924736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,51200,0.5538444519042969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,3072,64,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,16384,0.16557599438561335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,51200,0.25062399440341526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,65536,1.019301308525933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,12288,0.13222844070858425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,16384,0.08101599746280246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,16384,0.2474906709459093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,65536,1.0014844470553927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,51200,0.8004630936516656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,12288,0.06263911061816745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,12288,0.18115733729468453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,16384,0.23914665646023223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,51200,0.7912586530049642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,8192,0.10275289085176255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,10240,0.05390400025579664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,12288,0.19540799988640678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,10240,0.17100177870856392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,7168,0.08932799763149685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,10240,0.16460443867577448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,8192,0.042931556701660156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,8192,0.11629688739776611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,6144,0.07245866457621257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,7168,0.03762400150299072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,8192,0.1382382180955675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,5120,0.06206755505667793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,7168,0.10437333583831787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,6144,0.03557688991228739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,6144,0.09036266803741455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,7168,0.11348444885677761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,4096,0.04907022251023186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,5120,0.029290666182835896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,6144,0.09897955258687337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,3584,0.03855289022127787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,5120,0.0757582253879971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,4096,0.02497333288192749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,3072,0.03325422273741828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,4096,0.06109778086344401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,3584,0.022799111074871484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,3584,0.054324444797303945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,10240,0.12542488839891222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,2560,0.03048444456524319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,3584,0.06305066744486491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,3072,0.020577778418858845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,3072,0.04841422372394138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,2048,0.023452444208992854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,3072,0.056631108125050865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,2560,0.018814222680197824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,1536,0.01902044481701321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,2560,0.04158044523662991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,2560,0.049061334795422025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,2048,0.016812443733215332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,2048,0.03545066714286804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,1024,0.014074666632546319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,2048,0.041047109497918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,1536,0.014837332897716098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,1536,0.028378665447235107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,768,0.011437333292431302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,1536,0.03407111101680332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,1024,0.013072888884279462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,1024,0.02159022291501363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,512,0.008896888958083259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,1024,0.026891556051042344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,768,0.012101333174440594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,768,0.017903111047214933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,5120,0.08318133486641778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,256,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,768,0.023299554983774822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,4096,0.06894844770431519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,512,0.011185777684052786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,128,0.006069333189063602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,256,0.01070844464831882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,512,0.019425777925385367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,64,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,256,0.01274577776590983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2560,32,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,256,0.01706133286158244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2560,128,0.010954666468832227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,128,0.012156444291273752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2560,128,0.01349688900841607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,64,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,32,0.012372444073359171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,65536,0.5864951345655653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,65536,0.29941778712802464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,51200,0.4637609057956272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,16384,0.1598951154285007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2560,512,0.014847111370828418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,65536,0.8338951004876031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,51200,0.6542355749342177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,16384,0.07889155546824138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,65536,0.9267324871487088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,12288,0.12025154961480035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,16384,0.19877421855926514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,10240,0.10719555616378784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,12288,0.06001600291993883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,51200,0.7391573588053385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,12288,0.14963732825385198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,10240,0.051795555485619434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,8192,0.08546577559577094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,10240,0.12427643934885661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,7168,0.07056978013780382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,8192,0.04195911023351881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,8192,0.10052355792787339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,8192,0.12535199854109022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,7168,0.036107556687461004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,51200,0.2398515542348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,7168,0.08804355727301703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,6144,0.06392533249325223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,5120,0.0512640012635125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,7168,0.11363289091322158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,6144,0.03130577670203315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,6144,0.07674844397438897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,4096,0.04261599977811178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,5120,0.0650177796681722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,16384,0.23830578062269422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,5120,0.08198222186830309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,3584,0.03718044360478719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,4096,0.023339556323157415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,12288,0.18307733535766602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,4096,0.05304177933269077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,3072,0.03372088736957974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,4096,0.06826666990915935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,3584,0.02137600051032172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,10240,0.15494667159186468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,3584,0.0473911099963718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,2560,0.028036445379257202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,3584,0.06120444668663872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,3072,0.019455111689037748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,3072,0.04184800055291918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,2048,0.023592000206311543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,3072,0.05407377746370104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,2560,0.017879999346203275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,2560,0.0366542206870185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,1536,0.019409777389632333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,2560,0.04706933432155185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,2048,0.01593866613176134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,1024,0.014122666584120857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,2048,0.030393779277801514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,2048,0.03962666789690653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,1536,0.014540443817774454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,768,0.011020444333553314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,1536,0.024482667446136475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,5120,0.027428444888856676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,1536,0.0328453348742591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,1024,0.012448888685968188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,1024,0.018623110320832994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,512,0.009005332986513773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,768,0.01204711116022534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,1024,0.02550044490231408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,256,0.006646222124497096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,768,0.01588622232278188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,512,0.010947555303573608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,768,0.021958221991856892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,512,0.013171555267439948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,128,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,512,0.01887822151184082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,256,0.01036622209681405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,256,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,64,0.006304889089531369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,256,0.015573332707087198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,2048,32,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,2048,128,0.010462221999963125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,128,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,128,0.012083555261294046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,64,0.010724444356229572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,2048,32,0.010999999940395355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,65536,0.45033423105875653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,65536,0.24573066499498156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,2048,6144,0.09529955519570245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,51200,0.3788924482133653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,16384,0.12538222471872965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,51200,0.19385243786705864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,65536,0.6634817653232151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,51200,0.5157804489135742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,12288,0.0930711097187466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,16384,0.1626640028423733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,16384,0.06675022178226046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,65536,0.7677431106567383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,10240,0.08595999744203354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,12288,0.052630222505993314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,16384,0.20128088527255586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,8192,0.06398933463626437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,12288,0.12315377924177383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,51200,0.6072977913750542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,7168,0.054437332683139376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,10240,0.04439822170469496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,12288,0.154947559038798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,10240,0.10480444298850165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,8192,0.03363200028737386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,6144,0.047071999973720975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,8192,0.08363999923070271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,10240,0.13049866093529597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,7168,0.03016977839999729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,5120,0.03796089026663039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,8192,0.10718488693237305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,7168,0.07400977611541748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,6144,0.026899556318918865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,4096,0.028635554843478735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,7168,0.09432088666492039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,6144,0.06419377856784396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,3584,0.024460444847742718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,5120,0.02351911034848955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,5120,0.054422222905688815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,4096,0.019938665959570143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,3072,0.023226666781637404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,4096,0.0449848903550042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,3584,0.0185208883550432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,4096,0.05931911203596327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,3584,0.04056355688307021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,2560,0.019409777389632333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,3584,0.05297244588534037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,3072,0.017204445269372728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,2048,0.015863999724388123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,3072,0.035475555393430926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,2560,0.015665777855449252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,2560,0.030958222018347845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,1536,0.012805333567990197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,2560,0.04106222258673774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,2048,0.014210666219393412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,2048,0.026081777281231348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,1024,0.009476444787449306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,2048,0.03509777784347534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,1536,0.012830221818553077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,1536,0.021539555655585393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,768,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,1536,0.028781334559122723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,1024,0.011410666836632622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,1024,0.016186666157510545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,512,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,1024,0.023010666171709698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,768,0.010728889041476779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,768,0.014180445008807711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,5120,0.07051022185219659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,256,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,768,0.019919999771647982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,512,0.011744000017642975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,512,0.010047111246320937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,128,0.004839999808205498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,512,0.016861332787407767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,256,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,64,0.0042444442709287005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,256,0.009932444327407414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1536,32,0.004278222305907143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1536,128,0.009672000176376766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,3072,0.047573334640926786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,128,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,128,0.011984888878133563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,64,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1536,32,0.009318222602208456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,65536,0.3100915484958225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,65536,0.19564444488949248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,51200,0.2350755532582601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,6144,0.0821031133333842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,65536,0.5175671047634548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,51200,0.15402666727701822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,16384,0.08628533283869426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,51200,0.4028168784247504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,65536,0.6279608938429091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,12288,0.06580355432298449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,16384,0.05329066514968872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,16384,0.1331119934717814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,51200,0.49256356557210285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1536,256,0.0151591118839052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,10240,0.05346044566896227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,16384,0.16624622874789768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,12288,0.10002222326066758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,8192,0.04228533307711283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,10240,0.03712000118361579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,12288,0.12838133176167807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,10240,0.08414577775531346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,7168,0.03756177756521437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,8192,0.027087999714745417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,10240,0.10890044768651326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,8192,0.06716977887683444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,6144,0.03135200010405646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,8192,0.0890257755915324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,7168,0.02462933295302921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,7168,0.05951733059353299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,5120,0.026850667264726426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,7168,0.07861333423190646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,6144,0.02202577723397149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,4096,0.020765332712067496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,6144,0.06858933634228177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,5120,0.01937866707642873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,5120,0.044171555174721606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,3584,0.019001776973406475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,5120,0.05894666910171509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,4096,0.01681866745154063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,4096,0.03685599896642897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,3072,0.016163556112183463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,3584,0.01574577722284529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,12288,0.042528887589772545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,4096,0.049504889382256397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,2560,0.014206222361988492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,3584,0.04457955559094747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,3072,0.014490667316648694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,3072,0.028925332758161757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,2048,0.012262221839692859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,2560,0.025614221890767414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,2560,0.013426666458447775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,3072,0.039493332306543984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,1536,0.010619555910428366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,2560,0.03365244468053182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,2048,0.012182222472296821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,2048,0.02161688937081231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,1024,0.008358221915033128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,2048,0.029529776838090684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,1536,0.01106844428512785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,1536,0.01771022213829888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,768,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,1536,0.02444533341460758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,6144,0.05154222249984741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,1024,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,512,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,1024,0.013624000052611033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,768,0.009829333258999718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,1024,0.019911999503771465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,256,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,768,0.011703111231327057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,512,0.009367999931176504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,768,0.017514665921529133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,128,0.003986666599909465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,512,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,256,0.00867911097076204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,512,0.015216888652907478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,64,0.003556444413132138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,256,0.008672888908121321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,1024,128,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,1024,32,0.005196444276306364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,3584,0.03312266535229153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,256,0.013359111216333179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,128,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,1024,128,0.010710222025712332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,64,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,1024,32,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,65536,0.2555920018090142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,51200,0.19281689325968424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,16384,0.06481155421998766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,65536,0.19316889180077446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,51200,0.15254578325483534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,16384,0.05271910958819919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,65536,0.43622665935092503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,16384,0.11300088299645318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,51200,0.34173067410786945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,12288,0.04840266704559326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,10240,0.04021777709325155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,16384,0.16608977317810059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,12288,0.04163555635346307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,65536,0.6216693454318577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,51200,0.48735645082261825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,8192,0.031452443864610456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,12288,0.0855697790781657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,10240,0.07240355677074857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,10240,0.03678311241997613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,12288,0.12709777885013157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,7168,0.02549422283967336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,8192,0.026894221703211468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,10240,0.10795911153157552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,6144,0.0236453331179089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,8192,0.0579342246055603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,7168,0.024226667152510747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,5120,0.01962577799956004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,8192,0.08788977728949653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,7168,0.05187288920084635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,6144,0.021547555923461914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,7168,0.07750844293170504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,4096,0.01650399963061015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,6144,0.04490310947100321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,5120,0.019103111492262948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,6144,0.06771466467115614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,3584,0.014076444837782117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,5120,0.038520889149771795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,4096,0.01684711045689053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,5120,0.05796000030305651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,3072,0.012777778009573618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,4096,0.031706667608684964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,3584,0.015602666470739575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,4096,0.04865688747829861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,3584,0.028763555818133887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,2560,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,3072,0.014476444986131458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,3584,0.04343733191490173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,2048,0.009343110852771336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,3072,0.0255057778623369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,2560,0.013199110825856527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,1536,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,3072,0.03842844565709432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,2560,0.022086222966512043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,2048,0.01219377749496036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,2560,0.03300177719857957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,2048,0.018931556079122756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,1024,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,1536,0.01128266668981976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,768,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,1536,0.01575555569595761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,1536,0.024067555864651997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,1024,0.010267555713653564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,512,0.004585777719815572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,1024,0.01239822225438224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,1024,0.019371555911170114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,768,0.010572444233629437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,256,0.0038488888078265717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,768,0.017204445269372728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,512,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,512,0.01516355574131012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,128,0.003555555724435382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,256,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,256,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,256,0.01311822235584259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,64,0.0031555555760860443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,128,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,768,32,0.0032017777363459268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,128,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,128,0.010417777631017896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,64,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,65536,0.18954311476813424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,768,2048,0.02858577834235297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,51200,0.14611732959747314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,65536,0.14606222841474745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,768,0.009768000079525841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,65536,0.38420889112684464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,16384,0.04720888866318596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,768,512,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,51200,0.3023182286156548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,65536,0.47972620858086484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,51200,0.11673422654469807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,16384,0.04109511110517714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,16384,0.10089866982565986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,10240,0.030319998661677044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,51200,0.37772533628675675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,12288,0.03251377741495768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,16384,0.13018843862745497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,8192,0.02342222299840715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,12288,0.10139289167192246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,10240,0.027923554182052612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,10240,0.06382489204406738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,768,32,0.007154666715198093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,10240,0.08524888753890991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,8192,0.020115555988417733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,8192,0.052128000391854175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,6144,0.01871200071440803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,7168,0.018256889449225534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,7168,0.04597599969969856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,7168,0.061441779136657715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,5120,0.015961777832772996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,6144,0.04029511080847846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,6144,0.016552888684802584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,12288,0.03618311219745212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,4096,0.01369066701995002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,5120,0.014764444695578681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,5120,0.0344453321562873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,5120,0.046536889341142446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,12288,0.07634844382603963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,4096,0.02834222382969326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,3584,0.01167555567291048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,4096,0.013112000293201871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,4096,0.0381022228135003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,3072,0.010724444356229572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,7168,0.020947555700937908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,3584,0.012423999607563019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,3584,0.025405334101782903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,2560,0.009367110828558603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,3584,0.034514665603637695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,3072,0.011745777395036487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,8192,0.0694391131401062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,2048,0.008013333711359236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,3072,0.031097776359981958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,2560,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,2560,0.020224000016848247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,1536,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,2048,0.010393777655230628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,2048,0.01754044493039449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,6144,0.05384088887108696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,1024,0.005394666559166378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,2048,0.023707555400000677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,1536,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,1536,0.014152000347773233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,768,0.0045422220395671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,1536,0.019966221517986722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,1024,0.009376000199053023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,1024,0.011138666835096149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,512,0.004018666843573253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,1024,0.016562667157914903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,768,0.008986666798591614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,768,0.009734222458468543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,256,0.0032293332947625052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,768,0.0151235560576121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,512,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,512,0.008310221963458592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,128,0.0030951110853089225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,3072,0.02308888898955451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,256,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,256,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,64,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,512,32,0.002875555513633622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,512,128,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,128,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,2560,0.026925333672099646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,128,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,65536,0.11224355962541367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,64,0.0064586665895250105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,512,32,0.0063991112013657885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,51200,0.08747022019492255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,65536,0.14488266574011907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,65536,0.31016622649298775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,51200,0.11575555801391602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,16384,0.03404711021317376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,51200,0.2433697647518582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,12288,0.02609599961174859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,16384,0.04042311178313361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,65536,0.46277157465616864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,16384,0.08146933052274916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,51200,0.3638799985249837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,10240,0.02003466586271922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,16384,0.12679289446936712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,512,0.01309511148267322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,12288,0.06179733408821953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,8192,0.016781333420011733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,10240,0.028021332290437486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,512,256,0.011885333392355176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,10240,0.05185866687032911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,7168,0.015114666687117683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,10240,0.08249155680338542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,8192,0.04220444295141432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,6144,0.012938666674825879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,8192,0.0669742226600647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,7168,0.017900443739361234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,7168,0.03770044445991516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,5120,0.011305777562989129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,7168,0.05949422385957506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,6144,0.016361777981122334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,6144,0.03330844309594896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,4096,0.009336888790130615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,6144,0.05202310946252611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,5120,0.014727999766667684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,5120,0.0282008879714542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,3584,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,5120,0.04435111085573832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,4096,0.013069333301650153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,4096,0.02343911098109351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,3072,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,4096,0.036529776122834944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,12288,0.03217333224084642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,3584,0.012335999972290464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,3584,0.021316443880399067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,2560,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,12288,0.09779110882017349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,3584,0.033089776833852134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,3072,0.011729777687125735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,3072,0.019002666076024372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,2048,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,8192,0.01957244508796268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,2560,0.01678311162524753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,3072,0.029839999145931665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,1536,0.005195555587609609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,2560,0.02610933283964793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,2048,0.014481777946154276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,1024,0.0041848888827694785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,2048,0.010458666417333813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,2048,0.02296888828277588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,1536,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,768,0.003879111260175705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,1536,0.012088889049159156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,1024,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,1536,0.01927288870016734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,512,0.0031866667171319327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,1024,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,1024,0.015841777125994366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,768,0.00871288859181934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,768,0.00870400004916721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,256,0.003198222153716617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,768,0.014523555835088095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,512,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,512,0.007366221812036302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,128,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,512,0.01272266689274046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,256,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,256,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,64,0.0026177778426143858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,256,0.011049778097205691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,256,32,0.0028373334142896864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,128,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,65536,0.08238044712278578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,256,128,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,2560,0.011037333144081963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,64,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,256,32,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,51200,0.067413330078125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,65536,0.14427910910712347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,65536,0.26759910583496094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,16384,0.026729777455329895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,51200,0.11560266547732884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,51200,0.21039466063181558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,12288,0.018228444788191054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,16384,0.041313777367273964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,65536,0.4564702245924208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,16384,0.07044977611965604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,10240,0.015210666590266757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,51200,0.35891999138726133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,16384,0.12523110707600912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,12288,0.03182044294145372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,12288,0.05384622348679436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,8192,0.012958221965365939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,10240,0.02801688843303257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,256,128,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,12288,0.09690044323603313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,10240,0.045378665129343666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,7168,0.01146044416560067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,10240,0.08119377824995253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,6144,0.01015822258260515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,8192,0.0199635558658176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,8192,0.037010666396882795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,7168,0.019000000423855253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,5120,0.00867733359336853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,8192,0.06638755400975545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,7168,0.0328053335348765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,6144,0.016859556237856548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,4096,0.007256000406212277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,7168,0.05827199750476413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,6144,0.028821332587136164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,5120,0.014894222219785055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,3584,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,5120,0.02479377720091078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,6144,0.051495111650890775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,3072,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,4096,0.013190222283204397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,4096,0.020908445119857788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,3584,0.0127697777416971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,2560,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,4096,0.03649866580963135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,3584,0.01893422173129188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,3072,0.011726222104496427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,3584,0.03257955445183648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,2048,0.00463377767139011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,3072,0.01703466640578376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,2560,0.011056889262464313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,3072,0.029288887977600098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,1536,0.0042133331298828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,2560,0.015159999330838522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,2048,0.010728889041476779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,1024,0.0033253334048721525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,2048,0.013473778135246701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,1536,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,2048,0.02232622272438473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,1536,0.01108088923825158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,768,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,1536,0.019073777728610568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,1024,0.009016888837019602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,1024,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,512,0.002814222127199173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,1024,0.015522667103343539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,768,0.00868088917599784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,768,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,256,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,768,0.013791110780504016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,512,0.007115555306275685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,512,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,128,0.0021520000365045336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,5120,0.04369689027468363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,256,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,256,0.005728000154097875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,64,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,256,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,128,32,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3584,128,128,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,128,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,128,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,65536,0.07501688930723402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,64,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,51200,0.060772445466783315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,128,32,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,16384,0.02426311042573717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,2560,0.025847110483381484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,12288,0.014376888672510782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,10240,0.01237777786122428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,8192,0.013584000368913015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,65536,0.2680444452497694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,7168,0.010873777998818291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,16384,0.07016088565190633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,51200,0.2104017734527588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,12288,0.0535697771443261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,6144,0.009368889033794403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,10240,0.04518489042917887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,5120,0.008045333127180735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,4096,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,3584,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,8192,0.03660355673895942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,3072,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,6144,0.02863911125395033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,5120,0.02461244497034285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,2560,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,4096,0.020741333564122517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,2048,0.004519999855094486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,1536,0.0038684445122877755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,3584,0.018909333480728995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,1024,0.0032142222755485107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,3072,0.01685688893000285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,768,0.00286666676402092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,2560,0.014984889162911309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,2048,0.01311466677321328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,512,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,256,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,1536,0.01108088923825158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,128,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,1024,0.008950221869680617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,64,0.0018888889915413326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,768,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,64,32,0.0021128888345426982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,512,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,256,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,128,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,65536,0.07415378093719482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,64,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,16384,0.02017955482006073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,51200,0.059346669250064425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,32,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,12288,0.012368888490729861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,10240,0.010408889088365767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,8192,0.009218666288587783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,16384,0.07009689013163249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,65536,0.26838045650058323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,12288,0.053676442967520825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,7168,0.008845333423879411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,10240,0.04533066683345371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,51200,0.2102559937371148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3584,128,512,0.012655110822783576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,6144,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,5120,0.01015822258260515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,4096,0.008952000074916417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,8192,0.03659288750754462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,3584,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,3072,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,2560,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,7168,0.03256444467438592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,6144,0.0284586681260003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,2048,0.004569777597983678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,5120,0.024744888146718342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,4096,0.020808888806237113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,1536,0.0038737778862317405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,3584,0.019174221489164565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,3072,0.016932444439993966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,1024,0.0031964443624019623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,768,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,512,0.002502222235004107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,2560,0.015185778339703878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,256,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,128,0.0021395554973019496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,2048,0.01312711089849472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,1536,0.011132444772455426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,64,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,1024,0.009024000002278222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,768,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,512,0.007034666836261749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,256,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,128,0.005325333525737126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,64,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,32,32,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3584,64,7168,0.03237333231502109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3584,32,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,16384,1.4634808434380426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,12288,1.0927022298177083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,16384,3.9980489942762585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,12288,3.0544054243299694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,12288,3.430198245578342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,16384,4.781684451633029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,10240,0.8942542605929904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,8192,0.6922586758931478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,51200,5.535240173339844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,10240,3.2491857740614147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,10240,2.7807511223687063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,8192,2.3059874640570746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,8192,2.1617537604437933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,7168,0.67904265721639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,6144,0.5418311225043403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,7168,2.1290274726019964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,6144,1.694029384189182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,7168,1.9105146196153429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,6144,1.5632159974839952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,5120,0.46317598554823136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,4096,0.3754897647433811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,5120,1.3769324620564778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,5120,1.7105528513590496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,4096,1.1199501885308158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,4096,1.069350242614746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,3584,0.3377671241760254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,3584,1.066896014743381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,3072,0.30159465471903485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,3072,0.9414453506469727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,3584,0.9515066146850586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,2560,0.2761377758449978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,2560,0.7306453386942545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,3072,0.803800900777181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,2048,0.22156266371409097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,2560,0.7115964359707303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,2048,0.6341555383470323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,1536,0.49235910839504665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,2048,0.5493849118550619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,1536,0.17807555198669434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,1536,0.42868354585435653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,1024,0.28275465965270996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,768,0.21263110637664795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,1024,0.15949155224694148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,512,0.19016888406541613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,768,0.12204089429643418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,1024,0.4158053398132324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,256,0.13496355215708414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,768,0.2573475572797987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,128,0.10784800185097589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,512,0.2189404434627957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,256,0.09868266847398545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,256,0.19722667005327013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,64,0.07481155792872111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,32,0.07832266887029012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,128,0.11731733216179742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,128,0.18689421812693277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,64,0.19152977731492785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,65536,51200,13.351634555392794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,32,0.19237332873874238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,65536,512,0.12796089384290907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,16384,3.1680018107096353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,51200,4.169328053792317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,65536,5.602110120985244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,16384,1.1433031294080946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,65536,51200,20.02805413140191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,16384,3.7348124186197915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,12288,0.8538062307569715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,12288,2.846940358479818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,10240,2.0622230105929904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,51200,10.187370300292969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,12288,2.8280391693115234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,10240,0.6791253089904785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,10240,2.3156346215142145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,8192,0.548858642578125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,8192,1.664944012959798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,8192,1.697922706604004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,7168,0.48214043511284727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,7168,1.7466302447848852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,65536,12.642122904459635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,7168,1.440800878736708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,6144,0.4224960009256999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,6144,1.3659804662068684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,5120,0.3640302287207709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,6144,1.2172151141696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,4096,0.2973511219024658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,4096,0.9945128758748373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,5120,1.3186906178792317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,5120,1.0416044659084744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,4096,0.8354541990492078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,3584,0.7160346772935656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,3584,0.2865484555562337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,3584,0.7518924607170953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,3072,0.23690488603379992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,3072,0.7721920013427734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,2560,0.5229875776502821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,3072,0.6187617513868544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,2560,0.20744266774919298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,2048,0.5033840073479546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,2048,0.174782223171658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,1536,0.3235128985510932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,2048,0.4564968744913737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,1536,0.14002577463785806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,1024,0.24243201149834526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,1536,0.4310168955061171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,1024,0.11144088374243842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,768,0.1972986724641588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,1024,0.25984978675842285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,512,0.14022133085462782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,768,0.09744799799389309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,51200,15.719512939453125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,256,0.09333155552546184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,512,0.0975582202275594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,768,0.22236710124545625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,512,0.1724026600519816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,128,0.08075022035174899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,256,0.07791199949052599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,64,0.06189333067999946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,51200,32,0.06421422296100192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,51200,128,0.07666577895482381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,128,0.1507991154988607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,256,0.16247555944654676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,64,0.15004444122314453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,32,0.15250489446851942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,2560,0.576807975769043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,51200,1.3365181816948786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,16384,1.2755244572957356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,65536,1.893696043226454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,16384,0.39228089650472003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,12288,0.9554266399807401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,51200,3.313986672295464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,16384,1.3494880464341905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,12288,0.9107608795166016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,12288,0.3119893338945177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,10240,0.23270932833353677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,10240,0.7410017649332682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,51200,65536,20.891360812717014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,65536,4.455319298638238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,10240,0.7077093124389648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,8192,0.6138995488484701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,8192,0.21197333600785997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,7168,0.16781955295138887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,7168,0.5208871099683974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,8192,0.5652009116278754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,6144,0.1450373331705729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,7168,0.5015368991427952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,5120,0.39060621791415745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,51200,5.119396633572049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,5120,0.12485778331756592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,6144,0.4292453394995795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,4096,0.10274933444129096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,4096,0.2990906768374973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,3584,0.25904978646172416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,3584,0.09292178021536933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,4096,0.34458221329583066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,3584,0.24057955212063262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,3072,0.08130844434102376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,2560,0.19323110580444336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,3072,0.22295734617445204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,2048,0.1420631143781874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,2560,0.0716080003314548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,2560,0.178239107131958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,6144,0.48922046025594074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,1536,0.11956800354851617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,2048,0.06733689043256971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,1024,0.07811111211776733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,2048,0.15101867251926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,1536,0.05067466696103414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,65536,6.88793224758572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,768,0.0640488862991333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,1024,0.04253244400024414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,512,0.04787288771735298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,1536,0.11540177133348252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,5120,0.3493635654449463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,256,0.031163556708229914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,768,0.0402124457889133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,512,0.035184890031814575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,768,0.07043644454744127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,256,0.03152533372243246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,3072,0.23658490180969238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,512,0.06059822109010485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,64,0.023215111758973863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,16384,128,0.03124888737996419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,256,0.05423022309939066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,32,0.02312266661061181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,128,0.051401777399910815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,64,0.05220444334877861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,32,0.052857776482899986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,16384,1024,0.08839822477764553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,16384,128,0.028070221344629925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,51200,0.9539466434054904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,65536,1.289611604478624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,16384,0.7160284254286023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,16384,0.3433804512023926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,16384,0.8885004255506727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,12288,0.7982577747768826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,12288,0.208807110786438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,51200,2.4755511813693576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,10240,0.18750400013393828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,10240,0.5528373188442653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,12288,0.7862515449523926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,10240,0.5349902047051324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,8192,0.4012888802422418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,8192,0.14227022065056696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,65536,3.590579562717014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,7168,0.30366489622328013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,7168,0.1281342241499159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,8192,0.4865733252631293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,6144,0.33487288157145184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,7168,0.38793243302239316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,6144,0.11204177803463405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,5120,0.27155200640360516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,51200,3.982954661051432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,6144,0.32931023173862034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,5120,0.09802399741278754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,4096,0.2087235583199395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,4096,0.08708800209893121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,5120,0.3294808864593506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,3584,0.1843057738410102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,3072,0.17516444789038765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,3584,0.07154311074150933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,4096,0.25572000609503853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,2560,0.12063822481367324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,3584,0.1838488843705919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,3072,0.07059288687176175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,2560,0.056054221259223096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,2048,0.09832088814841376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,1536,0.08641866842905681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,3072,0.16539377636379665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,2048,0.04898755417929756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,2560,0.1374115546544393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,1024,0.06058400207095676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,1536,0.04083199964629279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,768,0.04500888784726461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,65536,5.1765725877549915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,2048,0.12361421849992539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,512,0.03385688861211141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,1024,0.0340266658200158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,256,0.024375110864639282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,1536,0.08950755331251357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,768,0.03144977821244134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,128,0.019448889626397025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,512,0.029166221618652344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,1024,0.06979999939600627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,768,0.05509688787990146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,64,0.01657511128319634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,256,0.025935999221271936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,12288,32,0.01623911162217458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,512,0.05065777897834778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,12288,128,0.025255999631351892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,256,0.0421653323703342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,128,0.040215111441082425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,64,0.04068888889418708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,12288,32,0.041703999042510986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,16384,0.6947795550028483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,16384,0.7419626447889539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,51200,0.8028915723164877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,16384,0.2381066746181912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,65536,1.0782995223999023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,12288,0.44300444920857746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,12288,0.181329780154758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,10240,0.14847644170125326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,10240,0.6327750947740343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,12288,0.6515084372626411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,10240,0.46880801518758136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,8192,0.34213688638475204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,8192,0.12170488304562038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,65536,2.634404500325521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,7168,0.2647911177741157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,7168,0.10901333226097955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,8192,0.5070924229092068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,6144,0.28442666265699595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,7168,0.3465511004130046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,5120,0.22340533468458387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,51200,3.115591049194336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,6144,0.27021243837144643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,5120,0.08050222529305352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,51200,2.017643610636393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,4096,0.18431911203596327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,5120,0.24466310607062447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,3584,0.1477031045489841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,4096,0.06787911388609144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,3584,0.06862044334411621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,3072,0.12825600306193033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,4096,0.20480533440907797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,3072,0.05462400118509928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,2560,0.11643377939860027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,3584,0.15627200073666042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,2048,0.08719644281599258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,2560,0.04977600110901726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,1536,0.06502044200897217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,2560,0.11852267053392197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,2048,0.04246222310596042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,1024,0.04501510990990532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,2048,0.09932533237669204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,1536,0.0351493325498369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,1536,0.07898310820261638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,768,0.038839999172422625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,65536,4.392180548773871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,6144,0.09557600153817071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,512,0.03278577658865187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,768,0.026142223013771906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,1024,0.058436446719699435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,256,0.01956266661485036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,512,0.029935999049080744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,768,0.047042667865753174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,128,0.017610667480362784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,512,0.039239999320771955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,256,0.021591110361946955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,64,0.014759999182489185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,10240,32,0.014658666319317289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,128,0.021200888686709937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,256,0.03428533342149522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,128,0.03278844555219015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,64,0.03274844421280755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,32,0.0325857765144772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,10240,3072,0.13637599680158827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,10240,1024,0.028502222564485338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,65536,0.8666266865200467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,16384,0.5602853563096788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,16384,0.1928328937954373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,51200,1.6392773522271051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,12288,0.4522799915737576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,65536,2.1533796522352433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,12288,0.15660799874199763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,16384,0.7750488917032877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,10240,0.3938737710316976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,12288,0.4378044340345595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,51200,0.6568649080064561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,10240,0.12394489182366265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,8192,0.32865333557128906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,8192,0.10079466634326512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,7168,0.2785084512498644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,10240,0.37808089786105686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,7168,0.11616533332400852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,51200,2.834391064114041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,8192,0.4823893441094293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,6144,0.08074933290481567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,6144,0.21360977490743002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,7168,0.2551315625508626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,5120,0.20784800582461885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,65536,3.3196230994330516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,4096,0.1571377780702379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,5120,0.06943821907043457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,4096,0.056497779157426625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,5120,0.22353598806593153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,3072,0.10642755693859524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,3584,0.051206221183141075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,4096,0.15355911519792345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,2560,0.08816444211535984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,3584,0.13556799623701307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,3072,0.04593866732385424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,2048,0.07765777905782063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,2560,0.04263110955556234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,1536,0.059608002503712974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,2560,0.09631466865539551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,2048,0.035347554418775774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,1024,0.039691554175482854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,2048,0.0800186660554674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,768,0.03256444467438592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,1536,0.06630666388405694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,1024,0.02353511088424259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,512,0.023471999499532912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,1024,0.048327998982535474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,768,0.02185955478085412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,256,0.016673778494199116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,768,0.03993955585691664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,512,0.019574221637513902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,6144,0.2262373235490587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,512,0.031664000617133245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,3584,0.1314986679289076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,256,0.018038221531444125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,64,0.012065778175989786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,256,0.027768888407283362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,128,0.017857778403494094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,32,0.012308444413873883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,128,0.02642755541536543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,64,0.026752000053723652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,3072,0.11219467057122125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,8192,32,0.026397332549095154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,8192,1536,0.028917332490285236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,8192,128,0.014363555444611443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,16384,0.4866008758544922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,65536,0.8966240353054471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,16384,0.1876711050669352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,51200,1.452789306640625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,12288,0.510484430525038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,16384,0.5216071340772841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,65536,1.932525422837999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,12288,0.14352177249060735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,12288,0.414048883650038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,10240,0.3818515671624078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,10240,0.11848355664147271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,8192,0.2299999925825331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,10240,0.31892532772488064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,8192,0.1142959992090861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,51200,0.669872866736518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,7168,0.19632800420125326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,8192,0.2586551242404514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,7168,0.09835466411378647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,6144,0.07311377922693889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,6144,0.19859200053744844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,7168,0.22239555252922905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,65536,2.9503805372450085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,5120,0.17043288548787436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,6144,0.19614756107330322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,4096,0.12356532944573297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,5120,0.0620702240202162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,3584,0.10378311077753703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,4096,0.058054222000969775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,5120,0.16201244460211858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,3584,0.04635111159748501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,3072,0.09130399756961399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,2560,0.0770053333706326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,3584,0.11424711015489365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,3072,0.04121066795455085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,2048,0.05949778027004666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,3072,0.10574311017990112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,2560,0.03648711244265238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,2560,0.08655110994974773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,1536,0.044907556639777295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,2048,0.031217777066760596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,1024,0.034119112624062434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,2048,0.07090844710667928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,51200,2.2378222147623696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,1536,0.02626666592227088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,1024,0.02197066611713833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,1536,0.056668447123633496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,512,0.01992177797688378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,1024,0.04340533415476481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,768,0.019699555304315355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,256,0.01444533301724328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,768,0.035579555564456515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,128,0.012366222010718452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,512,0.018409777018758986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,512,0.028767112228605483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,64,0.010463111102581024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,256,0.024999999337726172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,128,0.016346666547987197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,32,0.010998222563001843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,128,0.023990222149425085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,4096,0.14969600571526423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,64,0.024030221833123103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,7168,32,0.023687111006842718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,7168,768,0.028344889481862385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,65536,0.7118550936381022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,51200,0.5755555364820693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,7168,256,0.016512889001104567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,16384,0.4577324655320909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,16384,0.20112266805436876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,51200,1.201684421963162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,12288,0.33290667004055446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,65536,1.7221093707614477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,12288,0.12251910898420547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,12288,0.38262044058905703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,10240,0.2631768915388319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,10240,0.10161955489052667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,10240,0.2911653253767225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,8192,0.22835021548800996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,8192,0.08594133456548055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,7168,0.18213244279225668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,8192,0.22517244021097818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,7168,0.07437422540452746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,51200,2.068305757310655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,5120,0.13553778330485025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,7168,0.1962640020582411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,6144,0.06460622284147474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,65536,2.4994178348117404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,4096,0.10478578011194865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,6144,0.17132088873121473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,5120,0.05817244450251261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,5120,0.1574808888965183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,3584,0.0906755526860555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,3072,0.07809333006540935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,4096,0.11445777946048313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,3584,0.04187644521395365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,2560,0.06766311327616374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,3584,0.10195200310813056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,16384,0.5359048843383789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,2048,0.050254222419526845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,2560,0.03278577658865187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,2560,0.07680088943905301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,1536,0.0402746664153205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,2048,0.027942223681343928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,2048,0.06227199898825752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,1024,0.027327999472618103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,1536,0.02369955513212416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,6144,0.17264889346228707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,1024,0.01998844411638048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,1536,0.05043111244837443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,512,0.016689777374267578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,1024,0.03787466552522447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,768,0.01850844422976176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,768,0.03221688999070062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,512,0.016692444682121277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,256,0.012444444828563266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,4096,0.05004444387223986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,512,0.025194666451878015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,128,0.009988444546858469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,256,0.015322667029168872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,128,0.015453333655993143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,64,0.008770667016506195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,256,0.02199555602338579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,32,0.009694221946928237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,128,0.02115733259254032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,3072,0.08866133292516072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,64,0.021321778496106465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,6144,32,0.021130666136741638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,6144,768,0.023051554958025616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,65536,0.5993235376146104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,51200,0.8828408983018664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,51200,0.471910211775038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,6144,3072,0.038153777519861855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,16384,0.37292178471883136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,12288,0.23306666480170357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,16384,0.15067999892764622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,12288,0.12426311439938015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,10240,0.17977421813540986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,12288,0.28034310870700413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,16384,0.40758845541212296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,10240,0.09405599700080024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,8192,0.14510133531358507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,10240,0.25496355692545575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,51200,1.7474560207790797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,7168,0.15845510694715711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,8192,0.06987200180689494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,65536,2.020692401462131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,6144,0.13615288999345568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,8192,0.18832266330718994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,7168,0.06405333015653822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,5120,0.11504088507758246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,6144,0.055656890074412026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,7168,0.17314399613274467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,4096,0.09430133634143406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,65536,1.1409244537353516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,5120,0.04725155565473768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,3584,0.07810755570729573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,4096,0.03928088810708787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,3072,0.06279466549555461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,5120,0.13055733839670816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,2560,0.055141333076688975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,4096,0.10606222020255195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,3584,0.03517511155870225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,3072,0.031705776850382485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,2048,0.04472088813781738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,3584,0.08759199910693699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,1536,0.03570488757557339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,2560,0.02857066525353326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,2048,0.024615110622511968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,1024,0.023400889502631292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,2560,0.06499199734793769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,1536,0.022358222140206233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,2048,0.05429333448410034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,768,0.018852444158660043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,1536,0.04319733381271362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,512,0.014150222142537435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,1024,0.017881777551439073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,256,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,1024,0.03308799862861633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,512,0.014757333530320061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,768,0.0277502222193612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,128,0.009467555416954888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,256,0.013863999810483722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,512,0.021951110826598272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,64,0.008625778059164682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,256,0.01923733287387424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,5120,32,0.00869599978129069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,128,0.014088888963063559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,128,0.01846933364868164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,6144,0.1460231145222982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,64,0.018589332699775696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,32,0.018403554956118267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,5120,3072,0.07566310962041219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,16384,0.23776798778110078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,51200,0.37725422117445206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,5120,768,0.01627111103799608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,65536,1.008859528435601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,51200,0.8861733542548286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,16384,0.12246310710906982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,12288,0.18538667096032035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,16384,0.3205706543392605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,12288,0.11594843864440918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,12288,0.2733955648210314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,8192,0.16767911116282144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,51200,1.3191377851698134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,10240,0.19711821609073213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,65536,1.6738835440741644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,7168,0.13653955194685194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,8192,0.08103644185596041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,8192,0.16096533669365778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,6144,0.118212448226081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,7168,0.05701333284378052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,7168,0.13934132787916395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,65536,0.49646578894721133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,6144,0.04724977744950188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,4096,0.07396088706122504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,6144,0.1271324422624376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,5120,0.040186666780047946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,5120,0.10478666755888198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,3584,0.06332444482379489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,4096,0.03422222203678555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,3072,0.05145422286457486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,4096,0.0819280015097724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,10240,0.18800532817840576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,2560,0.04351910948753357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,3584,0.07322488890753852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,3072,0.027287999788920086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,10240,0.07505510913001166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,2048,0.035057776504092746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,3072,0.06429600053363375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,2560,0.024336000283559162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,1536,0.02831288841035631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,2560,0.05494399865468343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,2048,0.021012443635198805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,1024,0.02032533288002014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,1536,0.018239999810854595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,2048,0.04564977685610453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,768,0.016514667206340365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,1024,0.01536177761024899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,5120,0.10074844625261094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,512,0.012416889270146688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,1024,0.027855111493004694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,768,0.01439377831088172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,256,0.008897778060701158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,768,0.023629332582155865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,512,0.013442666994200813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,128,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,256,0.012728888955381183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,512,0.01880088945229848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,64,0.006561777657932705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,4096,32,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,256,0.016491555505328707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,128,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,4096,3584,0.03037066592110528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,128,0.015878222054905362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,64,0.015846222639083862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,32,0.015815999772813585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,65536,0.4384293291303847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,4096,1536,0.03705688979890611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,16384,0.2086391051610311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,51200,0.8143031332227918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,65536,0.9627626207139758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,12288,0.15088710520002577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,10240,0.14524711502922907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,12288,0.10099022256003486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,16384,0.31069421768188477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,8192,0.1272746721903483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,10240,0.08839378092024062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,65536,1.4487999810112848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,10240,0.1860355536142985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,7168,0.10923022031784058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,8192,0.07464088996251424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,8192,0.14244266351064047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,6144,0.1035226649708218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,7168,0.05118933320045471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,51200,0.3380088806152344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,5120,0.07404088973999023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,7168,0.12634132968054876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,6144,0.044052445226245456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,4096,0.05971022446950277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,6144,0.1086133321126302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,5120,0.037406222687827215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,16384,0.10343999995125665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,3584,0.0581173300743103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,5120,0.0914622214105394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,4096,0.0310533344745636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,3072,0.05043111244837443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,3584,0.02751733362674713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,2560,0.03915822174814012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,3584,0.06667821937137179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,3072,0.02535377773973677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,2048,0.03345511025852627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,51200,1.1518399980333116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,2560,0.022639110684394836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,1536,0.02712533374627431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,2560,0.049807111422220864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,2048,0.020033778415785897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,2048,0.042169776227739125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,1536,0.017516444126764934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,768,0.013527111046844058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,1536,0.03370133373472426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,1024,0.015042665931913586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,1024,0.025237333443429735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,512,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,768,0.013787555197874704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,256,0.007925333248244392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,768,0.02145333256986406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,512,0.012797333300113678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,128,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,512,0.017403556240929496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,256,0.011928888658682505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,256,0.0151235560576121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,64,0.005969777703285217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,4096,0.07570222351286146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,32,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,128,0.014490667316648694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,64,0.014561777313550314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,12288,0.30209332042270237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,3072,0.05823644664552477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3584,32,0.014503111441930136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3584,1024,0.017398221625222098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,65536,0.3723822169833713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,65536,0.7251280148824056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,51200,0.30771999888949925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,51200,0.5950791041056315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,16384,0.23518133163452148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,12288,0.17496800422668457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,16384,0.09576088852352566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3584,128,0.0120319997270902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,10240,0.14630844857957628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,16384,0.2629200087653266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,12288,0.0724453330039978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,51200,0.9287164476182727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,8192,0.10123466783099705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,12288,0.20516888300577799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,10240,0.06291377544403076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,7168,0.08572089009814793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,8192,0.052545779281192355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,10240,0.15681866804758707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,6144,0.07294488615459867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,7168,0.04435111085573832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,5120,0.06555022133721246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,6144,0.03744355506367154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,7168,0.11051911115646362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,4096,0.05690133571624756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,5120,0.03300977746645609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,3584,0.047409776184293956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,5120,0.083697775999705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,4096,0.027111111415757075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,4096,0.06595110893249512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,3072,0.04035377833578322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,3584,0.024736889534526404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,2560,0.032531556155946516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,3072,0.022437334060668945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,3584,0.06063822242948744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,2048,0.027297778262032404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,3072,0.05221244361665514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,2560,0.020290666156344943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,1536,0.021013332737816706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,65536,1.1082080205281575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,2560,0.04486133323775398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,2048,0.017921777235137094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,1024,0.015186667442321777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,2048,0.037885334756639265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,768,0.011615110768212212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,1536,0.0155395550860299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,512,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,1024,0.01346577786737018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,1536,0.030121776792738173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,768,0.012839111189047495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,256,0.006894222150246303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,1024,0.02281422250800663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,768,0.019647111495335896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,512,0.01200799975130293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,256,0.011051555474599203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,6144,0.10277155372831558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,512,0.015816888875431485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,64,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,32,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,3072,128,0.0111191107167138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,256,0.013756444056828817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,128,0.01310400002532535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,64,0.012800888882742988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,32,0.013459555804729462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,65536,0.31674133406745064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,65536,0.5720177756415473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,51200,0.43044622739156085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,51200,0.2524355517493354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,16384,0.17531288994683158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,12288,0.12608799669477674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,16384,0.08050044377644856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,3072,8192,0.13026755385928684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,10240,0.10453333457310994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,16384,0.22213333182864717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,3072,128,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,12288,0.060843553808000356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,65536,0.9544142617119683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,51200,0.7448408868577746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,8192,0.07970755630069308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,7168,0.07377777496973674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,10240,0.05668533510631985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,8192,0.041491554843054876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,10240,0.13767466280195448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,6144,0.06722488668229845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,7168,0.037189334630966187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,5120,0.05649244123035007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,8192,0.10968177848392063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,4096,0.04518844352828132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,6144,0.033082667324278094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,7168,0.09860977861616348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,3584,0.03864533371395535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,5120,0.028355555401908025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,6144,0.08573777808083428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,4096,0.02432800001568264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,3072,0.033540444241629705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,5120,0.0735751125547621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,2560,0.02847555610868666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,4096,0.05743910868962606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,3072,0.020276443825827706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,3584,0.051967998345692955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,2048,0.023403555154800415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,3072,0.046001778708563916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,1536,0.020114666885799833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,2560,0.03944177760018243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,2048,0.016163556112183463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,2048,0.03340177734692892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,1536,0.014541332920392355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,1536,0.027179555760489568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,12288,0.17554488446977404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,768,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,1024,0.012599999705950419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,1024,0.020246222615242004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,512,0.009810667071077559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,768,0.011800000237094032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,256,0.00738844441043006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,768,0.01721244388156467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,512,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,128,0.005872888697518243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,512,0.014152000347773233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,256,0.010402667025725046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,64,0.00535022219022115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,256,0.012345777617560493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,3584,0.022376888328128394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,128,0.010497777826256223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,32,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,128,0.011723555624485016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2560,2560,0.018536888890796237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,64,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2560,32,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2560,1024,0.014430221584108142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,65536,0.28857421875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,65536,0.5577911270989312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,16384,0.12569155957963732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,51200,0.426346672905816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,51200,0.22929954528808594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,12288,0.09369333585103352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,16384,0.08523288700315688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,10240,0.09590755568610297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,12288,0.0707004467646281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,8192,0.07736177576912774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,65536,0.766774230533176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,16384,0.18221955829196504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,12288,0.14004355006747776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,51200,0.6024559868706597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,7168,0.06290844413969252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,10240,0.11601066589355469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,6144,0.05484177668889364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,8192,0.039064000050226845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,5120,0.04614488946066963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,7168,0.035071998834609985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,8192,0.102856887711419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,6144,0.03064799971050686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,7168,0.08286488718456693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,5120,0.027621333797772724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,3584,0.02901155418819851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,5120,0.061378664440578885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,4096,0.022833777798546687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,3072,0.02552355494764116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,4096,0.0506915549437205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,3584,0.020959110723601446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,2560,0.023261333505312603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,3072,0.0192586663696501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,3584,0.0452497767077552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,2048,0.018553778529167175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,3072,0.039685332112842135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,2560,0.01754577789041731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,1536,0.014462222655614218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,2560,0.03420711225933499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,2048,0.015849777393870883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,1024,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,2048,0.028933333026038274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,1536,0.014097778333557976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,768,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,1536,0.02363288899262746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,1024,0.012151999606026543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,512,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,10240,0.056828445858425565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,1024,0.017831110292010836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,768,0.011392888923486074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,256,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,128,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,768,0.015577778220176697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,512,0.010616000327799054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,64,0.004432888908518685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,256,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,512,0.012749333348539142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,32,0.004609777695602841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,2048,128,0.01000711073478063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,256,0.011053333679835001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,6144,0.07367110914654203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,128,0.010392888552612728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,64,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,2048,32,0.010384889112578498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,51200,0.2873315546247694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,65536,0.23954576916164824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,65536,0.4400773313310411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,16384,0.11332800653245713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,51200,0.18653066953023276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,12288,0.08311021990246244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,16384,0.07424000236723158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,51200,0.4744550916883681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,65536,0.6049582163492838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,10240,0.07871555619769625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,16384,0.1534880002339681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,12288,0.05033066537645128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,8192,0.06387733088599311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,12288,0.11456888251834446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,7168,0.053694221708509654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,10240,0.04497333367665609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,8192,0.03353510962592231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,6144,0.04536799920929802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,10240,0.09661333428488837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,7168,0.03142755561404758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,2048,4096,0.035424000687069364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,5120,0.03580088747872247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,7168,0.06884444422192044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,4096,0.028543111350801256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,6144,0.026721777187453374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,3584,0.025955556167496577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,5120,0.02289688918325636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,4096,0.019567999574873183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,3072,0.022857778602176245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,5120,0.05060799916585287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,3584,0.0181697772608863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,4096,0.041834665669335254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,2560,0.01924088928434584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,3584,0.0378115541405148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,3072,0.016795555750528973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,2048,0.01629333363638984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,3072,0.033430222007963396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,2560,0.015509333875444202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,1536,0.013421333498424955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,2560,0.028591109646691218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,2048,0.013984888792037964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,1024,0.009370666411187913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,2048,0.024551999237802293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,1536,0.012685333689053854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,768,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,1536,0.020276443825827706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,512,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,1024,0.015511110424995422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,768,0.010614222122563256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,256,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,768,0.013182222015327878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,8192,0.07860889037450154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,512,0.00980177770058314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,128,0.004523555437723796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,512,0.011397333608733283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,256,0.009364444348547194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,6144,0.06068533658981323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,256,0.009670221971140968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,128,0.009296889106432596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,32,0.0042408886882993905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,128,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,64,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1536,32,0.009008888569143083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,65536,0.2733164363437229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,65536,0.1867848899629381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,16384,0.07832888762156169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,51200,0.14958843919965956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,16384,0.05599821938408745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,65536,0.464080015818278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1536,1024,0.011055111057228513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,16384,0.12086221906873916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,10240,0.04826755656136406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,12288,0.04090044564670987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,8192,0.03842577669355605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,12288,0.09149510992897882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,10240,0.03534488876660665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,7168,0.030442665020624798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,10240,0.0766542222764757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1536,64,0.003920000046491623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,8192,0.02630399995379978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,6144,0.026240888569090102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,7168,0.02405599918630388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,8192,0.06156088908513387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,5120,0.02272355556488037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,6144,0.02158044444190131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,7168,0.05486577749252319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,4096,0.018840889135996502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,51200,0.20765422450171578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,5120,0.01905066602759891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,3584,0.01687733332316081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,5120,0.04031999905904134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,4096,0.016541333662139047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,3072,0.01497244503762987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,4096,0.03352444370587667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,3584,0.015511999527613321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,2560,0.012868444124857584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,3584,0.03070488903257582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,12288,0.05761866437064277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,2048,0.010904889139864178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,51200,0.3664266798231337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,2560,0.013112000293201871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,3072,0.026893332600593567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,1536,0.008884444832801819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,2048,0.011881777809725868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,1024,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,2048,0.020000888241661918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,1536,0.010823999842007955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,768,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,1536,0.016737777325842116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,1024,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,512,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,1024,0.01277511070171992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,768,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,256,0.003925333420435588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,768,0.011321778098742167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,512,0.008984000318580205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,128,0.0035146665241983202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,512,0.009696000152164036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,256,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,64,0.003176888864901331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,128,0.0086666668454806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,128,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,1024,32,0.0031946665710873077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,3072,0.014466666513019137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,64,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,65536,0.2563564512464735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,32,0.007583111524581909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,2560,0.02329155637158288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,51200,0.19301954905192056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,65536,0.18574133184221056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,16384,0.06408355633417766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,51200,0.14831466144985622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,65536,0.41372177335951066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,51200,0.32693155606587726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,16384,0.056411557727389865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,10240,0.0420702232254876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,1024,256,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,12288,0.04042933384577433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,12288,0.08177155256271362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,10240,0.036072889963785805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,10240,0.07093422280417548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,7168,0.028799109988742407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,8192,0.026791110634803772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,8192,0.05556178092956543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,6144,0.024367999699380662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,7168,0.023777777949968975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,5120,0.020990222692489624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,6144,0.021408889028761122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,7168,0.049175110128190785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,4096,0.017841777867741056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,6144,0.04297955499755012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,5120,0.01870044403605991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,3584,0.015824000040690105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,5120,0.03752355443106757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,12288,0.04968444506327311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,4096,0.016547555724779766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,3072,0.013183111117945777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,4096,0.030436442957984075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,3584,0.01535999940501319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,2560,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,16384,0.10870577891667683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,3072,0.01425066590309143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,1024,6144,0.04768800073199802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,3584,0.027504889501465693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,1536,0.008354666332403818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,3072,0.024550222688251074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,2560,0.01313688854376475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,8192,0.031606223848130964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,1024,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,2048,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,2560,0.021357332666714985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,768,0.005877333382765452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,1536,0.010850666297806634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,2048,0.01848266687658098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,1024,0.01000711073478063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,512,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,256,0.0038968887594011095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,1024,0.011965333587593503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,768,0.009538667069541084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,128,0.0034942221310403612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,768,0.01038755559258991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,256,0.008632000121805403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,64,0.0031795555518733132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,128,0.00867377801073922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,32,0.003220444545149803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,128,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,64,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,65536,0.1799457735485501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,32,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,768,2048,0.009908444351620143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,51200,0.14469867282443577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,65536,0.14459999402364096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,16384,0.045406222343444824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,65536,0.33827466434902614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,51200,0.1123466624153985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,1536,0.015213333898120456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,12288,0.034535109996795654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,16384,0.039480000734329224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,51200,0.266067557864719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,768,512,0.008991999758614434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,16384,0.0920213328467475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,12288,0.030939555830425684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,512,0.008962666822804345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,8192,0.022239110536045496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,768,256,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,10240,0.027623999449941847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,7168,0.01867377758026123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,8192,0.01940533353222741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,10240,0.05677066908942329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,6144,0.016657777958446078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,8192,0.04564177658822802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,7168,0.017831110292010836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,5120,0.014526221487257214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,6144,0.016007999579111736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,7168,0.04128177629576789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,4096,0.011895111037625207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,6144,0.03614933292071024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,5120,0.014555555250909595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,3584,0.010989333192507425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,5120,0.030693332354227703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,4096,0.012795555922720166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,3072,0.010103999740547603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,4096,0.02549422283967336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,3584,0.012127999630239276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,2560,0.00869511150651508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,3584,0.022847111026446026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,3072,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,2560,0.010757333702511258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,3072,0.020469332734743755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,1536,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,2560,0.01791466606987847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,10240,0.027276444766256545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,2048,0.010195555786291758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,1024,0.004519999855094486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,2048,0.015505777464972602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,12288,0.06730044550365873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,768,0.0038942222793896994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,1536,0.009418666362762451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,512,0.003495111233658261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,1536,0.013131555583741931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,256,0.0029013332807355454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,1024,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,512,0.008010666403505538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,768,0.009298666483826106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,128,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,256,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,64,0.0025333333760499954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,256,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,128,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,32,0.002566222308410539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,128,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,65536,0.10583200057347615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,64,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,32,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,512,2048,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,51200,0.08266844352086385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,16384,0.03261333372857835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,65536,0.13971377743615046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,65536,0.29285865359836155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,51200,0.11139110724131267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,12288,0.02111022174358368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,16384,0.03872266742918227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,51200,0.23143556382921007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,10240,0.01649688846535153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,1024,0.008976889153321585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,16384,0.07748622364468045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,8192,0.013827555709415011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,512,768,0.008616888688670265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,12288,0.058337774541642934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,7168,0.012258666257063547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,8192,0.01926933394538032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,512,512,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,6144,0.011036444041464063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,8192,0.04014222158326043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,7168,0.01757600075668759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,5120,0.009367999931176504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,7168,0.035926222801208496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,6144,0.01610844499535031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,4096,0.007885333564546373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,5120,0.014488000008794995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,6144,0.03169066707293192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,3584,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,5120,0.02703199949529436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,4096,0.01277599980433782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,4096,0.022514666120211285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,3584,0.012100444071822695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,2560,0.006118222243256039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,3584,0.020399999287393358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,3072,0.011412444214026133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,3072,0.01824977828396691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,2048,0.004968889057636261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,2560,0.010716444088353051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,1536,0.004310222135649787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,2560,0.01629688839117686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,2048,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,12288,0.030454221698972914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,2048,0.014143110977278816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,10240,0.02676800058947669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,768,0.003176888864901331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,1536,0.011778666741318174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,10240,0.04909777641296387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,1024,0.008727110922336578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,512,0.0028773333049482773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,1024,0.009626666704813639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,768,0.008611555728647444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,768,0.008352888955010308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,256,0.002500444443689452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,512,0.008006222546100616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,128,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,512,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,256,0.007693332930405934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,3072,0.006334222025341458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,64,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,32,0.002266666748457485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,128,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,128,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,65536,0.07954755756590101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,64,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,32,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,51200,0.062374220954047314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,65536,0.13784444332122803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,256,1024,0.003551111039188173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,16384,0.024319110645188227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,65536,0.25301512082417804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,256,1536,0.009580444130632613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,51200,0.11078844467798869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,51200,0.1981173356374105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,16384,0.039770666095945574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,10240,0.013545778062608508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,12288,0.03105777833196852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,8192,0.012821333275900947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,12288,0.05081599950790405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,10240,0.026942221654786006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,7168,0.011434666812419891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,10240,0.04275288846757677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,8192,0.019687111179033916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,6144,0.009914666414260864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,256,256,0.006078222145636876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,8192,0.034958223501841225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,7168,0.01817244456874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,5120,0.008533333738644918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,4096,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,7168,0.03103999959097968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,6144,0.027525333894623652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,5120,0.01481244464715322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,3584,0.00666222224632899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,5120,0.023528888821601868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,3072,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,3584,0.012129777835475074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,3584,0.018179555733998615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,2560,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,12288,0.01570044457912445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,3072,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,2048,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,16384,0.06652000215318468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,2560,0.014498665928840637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,1536,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,2048,0.0127697777416971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,1536,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,1024,0.0035004446076022256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,1536,0.010779555473062726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,6144,0.016843555702103507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,768,0.003160888950030009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,1024,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,4096,0.013076444466908773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,512,0.002806222273243798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,4096,0.019955555597941082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,768,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,256,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,512,0.008001777860853408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,512,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,3072,0.016499555773205228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,128,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,256,0.007669332954618666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,2560,0.011074666347768573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,64,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,128,32,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,128,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,2048,0.01036622209681405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,128,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,65536,0.07068889008627997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,64,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,32,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,16384,0.021697777840826247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,12288,0.013444444371594323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,65536,0.25269420941670734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,10240,0.011764444410800934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,1024,0.009017777939637503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,12288,0.05082222157054477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,8192,0.010151111417346531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,7168,0.009399111072222391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,10240,0.0427724454138014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,6144,0.009993777506881291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,7168,0.030863109562132094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,128,256,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,5120,0.008997333546479544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,4096,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,6144,0.027130666706297133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,3584,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,51200,0.05661244524849785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,5120,0.023362666368484497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,3072,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,4096,0.019932443896929424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,2560,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,3584,0.018066666192478605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,2048,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,51200,0.19803644551171196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,3072,0.016214221715927124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,1536,0.0038444445365005066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,16384,0.0660382244322035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,1024,0.0031991110493739447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,2560,0.014495111174053617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,768,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3329,128,768,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,512,0.002575111058023241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,256,0.0022293333378103045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,2048,0.012720888687504662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,1536,0.01087288889620039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,1024,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,64,0.002157333410448498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,8192,0.0347377790345086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,512,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,256,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,32,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,128,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,65536,0.07504000266393025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,64,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,51200,0.056256890296936035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,32,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,12288,0.011668444507651858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,65536,0.25268533494737416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,10240,0.010084444450007545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,51200,0.19794400533040366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,8192,0.008962666822804345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,12288,0.05051111181577047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,7168,0.009015999734401703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,10240,0.04275911053021749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,6144,0.008269333177142674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,8192,0.03456444541613261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,5120,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,64,128,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,7168,0.03084355592727661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,4096,0.008983111215962304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,3584,0.008119111259778341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,64,768,0.008260444634490544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,3072,0.007324444750944774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,6144,0.027132444911532935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,2560,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,5120,0.023378666904237535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,2048,0.004512888689835866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,4096,0.019768888751665752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,1536,0.0038399998512532976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,3584,0.0181697772608863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,1024,0.0031662223239739737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,768,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,3072,0.01647911138004727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,512,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,2560,0.014477334088749357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,256,0.0022053333620230355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,1536,0.010761777559916178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,1024,0.008815111385451423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,128,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,768,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,64,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,32,0.0018915554715527429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,512,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,256,0.005659555395444234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,128,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,64,0.005286222116814719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,32,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3329,32,16384,0.020602666669421725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,2048,0.012783110969596438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3329,32,16384,0.06613066461351183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,16384,1.389438205295139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,16384,4.382477230495877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,16384,4.640411376953125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,51200,5.2239299350314665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,12288,3.2282685173882375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,16384,4.0898284912109375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,12288,0.9954026540120443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,12288,3.32989501953125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,10240,2.8016870286729603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,10240,0.8313733206854926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,10240,2.689078225029839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,12288,3.009036382039388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,8192,0.6586737632751465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,8192,2.2551697625054254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,10240,2.508035659790039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,51200,12.780545552571615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,7168,2.008497873942057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,8192,2.2088292439778647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,7168,1.8090409172905817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,7168,0.5795653131273057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,8192,2.0179600185818143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,7168,1.7749982410007055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,6144,0.5033831066555446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,6144,1.6305049260457356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,6144,1.567954699198405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,5120,1.222741338941786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,5120,0.4461670981513129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,4096,1.017496003044976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,6144,1.5699289109971788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,5120,1.34716977013482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,4096,1.0372141732109916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,5120,1.1598008473714192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,4096,0.3668222162458632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,3584,0.8897831175062391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,3584,0.3274764484829373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,4096,0.9948000378078885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,3584,0.8898142708672417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,3072,0.7878986464606391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,3584,0.8505484263102213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,3072,0.8068764474656848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,3072,0.29044000307718915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,2560,0.6907111273871528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,2560,0.2540960046980116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,3072,0.7454177538553873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,2048,0.5646808942159017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,2560,0.7195244365268283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,51200,19.025227864583332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,2560,0.6060524516635472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,2048,0.21176444159613717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,1536,0.4797102080451117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,1536,0.17503378126356336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,1024,0.28466934627956814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,1536,0.42673601044548887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,51200,14.128283182779947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,2048,0.8000088797675239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,1536,0.39665865898132324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,768,0.19446578290727404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,1024,0.15304622385236952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,768,0.11862666077084011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,512,0.16117421785990396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,1024,0.38189244270324707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,1024,0.2979128890567356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,768,0.30588266584608287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,768,0.24314755863613555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,256,0.11594488885667588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,512,0.11445600456661648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,128,0.09187822209464179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,512,0.22301689783732095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,512,0.1956311066945394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,256,0.1941546731524997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,256,0.16032889154222277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,64,0.07435199949476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,128,0.09124800231721665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,65536,32,0.08204178015391032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,128,0.11825954914093018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,65536,2048,0.51473601659139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,64,0.18700267208947075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,32,0.18897599644131133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,65536,256,0.09286844730377197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,65536,128,0.18641332785288492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,51200,3.8836771647135415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,65536,5.548593733045791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,16384,3.844673580593533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,51200,10.016019185384115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,16384,4.0425669352213545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,16384,0.9771671295166016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,65536,13.113642374674479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,12288,2.725583182440864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,16384,3.2809261745876737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,12288,0.7634008725484213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,51200,14.909506903754341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,51200,11.083311292860243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,12288,2.666373358832465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,12288,2.3137423197428384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,10240,0.7042613559299045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,10240,2.5825439029269748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,10240,2.2457474602593317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,8192,1.6342168384128148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,10240,1.9750311109754775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,8192,0.5988337728712294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,8192,1.7090622584025066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,7168,0.45728090074327254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,7168,1.5230017768012152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,65536,14.064340379503038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,8192,1.5494409137301977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,7168,1.433741357591417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,6144,1.2863555484347873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,6144,0.42934489250183105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,65536,20.43364800347222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,7168,1.3166204028659396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,5120,1.0691377851698134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,6144,1.2752550972832573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,5120,0.3444311088985867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,5120,1.0079537497626412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,6144,1.163599967956543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,4096,0.2900133397844103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,4096,0.8316924307081434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,3584,0.26002221637301975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,3584,0.7289288838704427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,5120,0.928259531656901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,3584,0.7021146880255805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,3072,0.6335973209804958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,4096,0.7639128896925184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,3072,0.23474489317999947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,3584,0.686712000105116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,2560,0.4910275671217177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,3072,0.6350293159484863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,2560,0.20182577768961588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,3072,0.5758319960700141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,2048,0.4342648983001709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,2560,0.5923431184556749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,4096,0.851752863989936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,2048,0.16801155938042533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,2560,0.4788826836480035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,2048,0.4568826887342665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,1536,0.3484453360239665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,1536,0.13639911015828451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,1024,0.22581333584255645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,2048,0.4038115607367621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,1536,0.30562133259243435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,1024,0.26471734046936035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,1024,0.10773599810070461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,768,0.09512533081902398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,512,0.1331119934717814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,1024,0.342120885848999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,768,0.19221422407362196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,256,0.08929510911305745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,512,0.11117688814798991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,512,0.17002577251858184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,512,0.15961955653296575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,128,0.06961688730451795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,256,0.07540000147289701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,256,0.15250221888224283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,1536,0.33952977922227645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,64,0.05835821893480089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,256,0.12216711044311523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,768,0.17549778355492485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,128,0.14574755562676325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,51200,128,0.09338044457965428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,64,0.14739289548661974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,32,0.14965777926974827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,51200,768,0.2041591140958998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,51200,32,0.060666667090521924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,65536,1.5715138117472331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,51200,128,0.07369510995017158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,51200,3.215579562717014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,16384,0.9855360454983182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,51200,1.2405004501342773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,16384,0.4118684397803412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,16384,1.1789129045274522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,16384,1.038643519083659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,65536,4.110216776529948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,12288,0.7906124326917859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,12288,0.25853689511617023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,51200,5.789418538411458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,12288,0.8421493106418185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,51200,3.487069235907661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,65536,6.128111945258246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,65536,4.4556905958387585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,10240,0.6773724555969238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,10240,0.22244087855021158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,12288,0.7233040067884656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,8192,0.5775813526577419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,8192,0.1920559936099582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,10240,0.6852009031507703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,8192,0.5582533412509495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,7168,0.4611217710706923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,10240,0.5970871183607314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,8192,0.524129761589898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,7168,0.15514222780863443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,7168,0.49435822168986004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,6144,0.4465973112318251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,6144,0.13841688632965088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,5120,0.33436176511976456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,6144,0.40361955430772567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,7168,0.5108933448791504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,5120,0.13992710908253989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,4096,0.2499155468410916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,5120,0.34358400768703884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,5120,0.3127555582258436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,3584,0.24910667207505968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,4096,0.314517339070638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,4096,0.2504017882876926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,3584,0.2592471175723606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,3584,0.22013510598076713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,3072,0.07889866828918457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,2560,0.17047466172112358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,3072,0.21607644028133818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,2560,0.06849066416422527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,3072,0.1936195558971829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,2560,0.17760888735453287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,2048,0.14484088950686985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,2560,0.18790577517615425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,1536,0.10386666986677383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,4096,0.09676622019873725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,2048,0.059677335951063365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,2048,0.14077777332729763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,1536,0.049152887529797025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,3584,0.09049689107471043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,1024,0.07454933060540093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,1536,0.115337782435947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,768,0.051273776425255664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,1536,0.11297510729895698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,1024,0.04045688774850633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,1024,0.08764088816112942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,1024,0.08585155672497219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,512,0.04386666748258802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,768,0.03716266817516751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,768,0.06994933552212186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,768,0.07348799705505371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,256,0.028894222444958154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,512,0.03433510992262099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,6144,0.3900444507598877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,512,0.06012800004747179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,256,0.05313600103060404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,128,0.025751110580232408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,256,0.030466665824254353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,256,0.047356443272696606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,64,0.021405332618289526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,3072,0.2076808876461453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,16384,32,0.021597334080272253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,128,0.05058844553099739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,16384,128,0.03614933292071024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,64,0.051423999998304576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,32,0.05215733249982198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,512,0.05925777885648939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,16384,2048,0.14374755488501653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,16384,128,0.030076444149017334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,65536,1.221127086215549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,51200,0.9257688522338867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,16384,0.9468515184190539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,65536,3.078671137491862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,51200,2.539151085747613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,16384,0.25160267617967397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,12288,0.6224017673068577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,16384,0.9457439846462674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,16384,0.7688755459255643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,12288,0.6380382113986546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,51200,3.532236523098416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,10240,0.5167751312255859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,51200,2.7090702056884766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,65536,3.461536831325955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,10240,0.5176551077100966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,10240,0.19148266315460205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,10240,0.4580248726738824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,8192,0.4343546761406793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,65536,5.593919965955947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,8192,0.15555289056566027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,7168,0.33423378732469344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,8192,0.5463768641153971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,12288,0.19621066252390543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,7168,0.12092888355255127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,8192,0.3591262234581842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,6144,0.29589777522616917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,7168,0.43479288948906797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,7168,0.33109066221449107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,6144,0.10644977622561985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,5120,0.24851645363701713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,12288,0.5809146563212076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,6144,0.31891732745700413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,6144,0.27655378977457684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,4096,0.198089771800571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,4096,0.07580088906817965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,3584,0.17742844422658285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,5120,0.30627555317348903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,5120,0.2377155621846517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,4096,0.2647555669148763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,4096,0.19502311282687715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,3072,0.14612000518374973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,3584,0.06785866949293348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,3584,0.18996711572011313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,2560,0.12203555636935765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,3072,0.06185422341028849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,3072,0.1582453383339776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,3584,0.17324444982740614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,2048,0.10200533601972793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,2560,0.056087109777662486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,2560,0.13850043879614934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,1536,0.08163466718461779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,2560,0.13171199957529703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,2048,0.0469742218653361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,1024,0.053655998574362866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,2048,0.11158488856421576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,1536,0.039714667532179095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,2048,0.1114835606680976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,1536,0.08833066622416179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,768,0.044255110952589244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,5120,0.09424355294969346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,1024,0.033082667324278094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,1536,0.08831377824147542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,512,0.03225244416130914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,768,0.03018222252527873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,1024,0.06879999902513292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,768,0.05433333251211378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,768,0.05901155869166056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,512,0.027973333994547527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,512,0.04632444514168633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,128,0.01976444489426083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,512,0.049350221951802574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,256,0.025263999899228413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,256,0.04165688819355435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,64,0.016674665941132438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,256,0.038479109605153404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,3072,0.1605564488304986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,12288,128,0.024375110864639282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,32,0.017318222257826064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,128,0.039348443349202476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,12288,128,0.029508444997999404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,64,0.039926220973332725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,32,0.040687110688951277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,12288,1024,0.06815289126502143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,12288,256,0.022659555077552795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,65536,0.9478658040364584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,51200,2.157343970404731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,65536,2.6478203667534723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,51200,0.7647280163235135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,16384,0.6846915351019965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,16384,0.24899466832478842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,16384,0.7225146823459201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,51200,3.046240064832899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,12288,0.4595644209120009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,16384,0.6616177558898926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,65536,3.969801584879557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,51200,2.3228585984971786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,12288,0.6103208859761556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,65536,2.952092488606771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,10240,0.48934846454196507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,12288,0.47597424189249676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,10240,0.14758933915032282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,8192,0.3325608836279975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,10240,0.4805022345648871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,8192,0.11588266160753037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,7168,0.29436977704366046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,10240,0.4013040065765381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,8192,0.39100178082784015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,7168,0.10326488812764485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,8192,0.35956533749898273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,6144,0.2486168808407254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,7168,0.34134933683607316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,7168,0.28323377503289116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,6144,0.09102400143941243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,5120,0.22376532024807402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,6144,0.2751573456658257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,12288,0.1691297822528415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,5120,0.07887288596895006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,4096,0.16729156176249185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,6144,0.2456728882259793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,5120,0.23413422372606066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,3584,0.1603813303841485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,5120,0.21249155203501383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,4096,0.19199732939402261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,3584,0.059205333391825356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,3072,0.12559021843804255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,4096,0.17019732793172201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,3584,0.1532542175716824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,2560,0.10737244288126628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,3072,0.05247288942337036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,3584,0.1513706710603502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,3072,0.13382221592797175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,2048,0.08715289168887669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,2560,0.05151999990145365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,3072,0.13681244850158691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,2560,0.1155724392996894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,1536,0.06717689169777764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,2048,0.040469331873787776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,2560,0.11446844206915961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,2048,0.09512888722949558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,1024,0.04842666784922282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,1536,0.035169776942994856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,2048,0.09731911288367377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,1536,0.07579111390643649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,768,0.03983022107018365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,1024,0.02791022260983785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,1536,0.07958400249481201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,512,0.028168890211317275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,1024,0.057598219977484814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,768,0.026258667310078938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,1024,0.06156444549560547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,256,0.026720888084835474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,768,0.04646933409902784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,512,0.0236462222205268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,768,0.052489777406056724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,512,0.03825777769088745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,128,0.017495110630989075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,256,0.020909332566791113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,512,0.04323911004596286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,64,0.01477688882086012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,256,0.034156444999906756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,10240,32,0.014800000521871777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,128,0.020450666546821594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,256,0.034336888127856784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,128,0.03231022093031142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,10240,128,0.025913778278562758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,64,0.03236888845761617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,10240,32,0.032058666149775185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,10240,4096,0.06641422377692328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,65536,0.793698681725396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,51200,1.6058097415500219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,16384,0.6713644663492838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,65536,2.008811526828342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,51200,0.6464471287197536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,16384,0.18449600537618002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,12288,0.5060470898946127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,16384,0.6076649030049642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,51200,2.597015168931749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,51200,1.9036301506890192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,12288,0.1395217842525906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,16384,0.5931324428982204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,12288,0.4399440023634169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,65536,3.067469278971354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,10240,0.36411910586886936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,10240,0.11915466520521377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,12288,0.4105937745836046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,8192,0.2801644537183973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,10240,0.38509511947631836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,8192,0.09802489148245917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,7168,0.24664976861741808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,10240,0.3358168866899278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,8192,0.3695057762993707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,8192,0.26604265636867946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,7168,0.2516062259674072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,7168,0.23677065637376574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,6144,0.0757733318540785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,6144,0.22280976507398817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,5120,0.1704204479853312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,6144,0.20393421914842394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,65536,2.5007076263427734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,5120,0.07092710998323229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,5120,0.177748441696167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,4096,0.13888000117407903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,5120,0.1741013394461738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,3584,0.11516355143653022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,4096,0.05445333321889242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,4096,0.14344799518585205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,3584,0.048968000544442065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,4096,0.1469635566075643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,3072,0.09840444723765056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,6144,0.2089466651280721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,3584,0.12667911582522923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,3584,0.139408893055386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,2560,0.08067022429572211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,3072,0.044489777750439115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,3072,0.11051644219292535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,2048,0.0670293304655287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,2560,0.03923377725813124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,3072,0.11338667074839275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,2560,0.09511288669374253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,1536,0.05583466423882378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,2560,0.09823111030790542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,2048,0.03426933288574219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,1024,0.03878044419818454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,2048,0.07903289132648043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,1536,0.027659555276234944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,2048,0.08206399944093493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,1536,0.06224977970123291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,1536,0.06672888994216919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,1024,0.023016000787417095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,1024,0.04795822170045641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,512,0.02055644492308299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,1024,0.0522311098045773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,768,0.021108445194032457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,768,0.0389342208703359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,256,0.015261333849694995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,768,0.04485511117511325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,512,0.019633778267436557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,512,0.031104889180925157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,512,0.03747111227777269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,256,0.017439999514155917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,256,0.027348443865776062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,64,0.011683555940786997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,256,0.029430223835839167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,32,0.012107555237081317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,128,0.026288888520664636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,8192,128,0.02272355556488037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,64,0.026292444931136236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,8192,32,0.026418666044871014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,768,0.028027554353078205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,7168,0.09530933035744561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,8192,128,0.017325333423084684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,65536,0.7157706684536405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,65536,2.0202354855007596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,51200,1.576491567823622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,51200,0.545721795823839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,8192,128,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,16384,0.5213946766323513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,16384,0.16067643960316977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,16384,0.5388888782925075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,12288,0.40730399555630153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,65536,2.957823011610243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,65536,2.1327431996663413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,51200,2.116037368774414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,12288,0.3790648778279622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,12288,0.12830311722225612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,10240,0.30161245663960773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,10240,0.10490755240122478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,12288,0.34829778141445583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,51200,1.6730995178222656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,8192,0.24571288956536186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,10240,0.3729537857903375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,8192,0.08591466479831272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,7168,0.21026399400499132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,8192,0.29017066955566406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,6144,0.16989600658416748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,8192,0.23152621587117514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,7168,0.2311697800954183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,7168,0.21270488368140328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,6144,0.06688355737262301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,5120,0.15047021706899008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,6144,0.1947395536634657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,6144,0.1925742228825887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,5120,0.05802222092946371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,4096,0.1118497716055976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,5120,0.16011022196875677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,5120,0.15458222230275473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,4096,0.04859911070929634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,3584,0.09954844580756293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,4096,0.12800089518229166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,10240,0.2961946593390571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,3072,0.08904000123341878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,4096,0.13153244389428034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,3584,0.04330577784114414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,7168,0.07692178090413411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,3584,0.11267377270592584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,2560,0.0757217804590861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,3072,0.03879822293917338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,16384,0.5125173462761773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,3072,0.09839289055930243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,3072,0.1020231114493476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,2048,0.06586577494939168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,2560,0.034783111678229436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,2560,0.0851973361439175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,1536,0.045942223734325834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,2560,0.08559111091825698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,2048,0.02977422210905287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,1024,0.03361422154638503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,2048,0.06992355320188734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,1536,0.05565155877007378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,2048,0.07426400317086114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,768,0.024131556351979572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,1536,0.06061955293019613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,1024,0.020993777447276648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,1024,0.04257955484920078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,512,0.018404444058736164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,1024,0.04639110962549845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,768,0.018790221876568265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,256,0.013037333057986366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,512,0.017133333616786532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,3584,0.11450310548146565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,512,0.028183112541834515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,128,0.01072177787621816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,512,0.03347377644644843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,256,0.015757333901193406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,256,0.024472888973024156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,64,0.009675555759006077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,256,0.025948445002237957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,128,0.01607022186120351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,7168,32,0.010006222460005019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,128,0.0207342223988639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,64,0.023549333214759827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,7168,1536,0.025177778469191656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,32,0.023357333408461675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,768,0.03519999980926514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,7168,768,0.04059733284844293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,65536,0.7001982265048556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,7168,128,0.02367288867632548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,51200,1.2557030783759224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,16384,0.35303555594550234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,65536,1.763152864244249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,51200,0.5811262130737305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,16384,0.1741795539855957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,16384,0.5047270986768935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,12288,0.3439822196960449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,16384,0.4260462125142415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,65536,2.3520560794406467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,51200,1.797790315416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,12288,0.34201778305901426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,10240,0.24158665868971083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,51200,1.4359893798828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,12288,0.14616799354553223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,10240,0.09663110971450806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,12288,0.31128888660007054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,8192,0.22005155351426867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,10240,0.2832533253563775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,10240,0.272544887330797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,7168,0.19786222775777182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,8192,0.2400888866848416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,8192,0.2666551007164849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,6144,0.16795378261142305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,7168,0.22371376885308158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,7168,0.1881413327323066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,6144,0.06598489152060615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,5120,0.13514399528503418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,6144,0.1687084436416626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,4096,0.10654399792353313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,6144,0.1662044392691718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,5120,0.053027557002173535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,5120,0.13935466607411703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,65536,1.8602169884575739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,4096,0.04401333464516533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,5120,0.1423226727379693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,3584,0.09097422493828668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,4096,0.11207289165920681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,3072,0.08499822351667617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,4096,0.11510577466752793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,3584,0.03977866636382209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,3584,0.09916177723142837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,8192,0.07764355341593425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,2560,0.06757689184612699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,3072,0.03457866774664985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,3584,0.1024106674724155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,3072,0.08695555395550197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,2048,0.05894133117463854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,7168,0.06932000319163005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,3072,0.09069866604275173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,2560,0.030770666069454614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,2560,0.07464533381991915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,1536,0.0421653323703342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,2048,0.026890666948424444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,2048,0.062246223290761314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,1536,0.02238399949338701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,2048,0.06668000088797675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,1536,0.05525333351559109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,768,0.022023111581802368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,1024,0.018583110637134977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,1024,0.03783644570244683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,1024,0.04288622074657016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,512,0.016528889536857605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,768,0.017151110702090792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,768,0.03159200151761373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,768,0.03689600030581156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,256,0.012757333616415659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,512,0.015844444433848064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,512,0.03134311238924662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,256,0.021670222282409668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,256,0.01479555500878228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,128,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,2560,0.07986399862501356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,256,0.023576888773176406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,1024,0.02920177910063002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,64,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,6144,32,0.00870666652917862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,6144,128,0.014679110712475248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,128,0.020976000361972388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,6144,128,0.01869155632125007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,64,0.020925333102544148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,32,0.02067377832200792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,512,0.025193777349260118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,65536,0.5628702375623915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,51200,0.8963901731703016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,6144,1536,0.049366222487555615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,65536,1.3566435707939997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,16384,0.34600088331434464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,51200,0.4389804469214545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,16384,0.37524623341030544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,16384,0.12796799341837564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,16384,0.40627288818359375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,12288,0.29411289427015513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,65536,2.0964391496446395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,51200,1.6207093132866754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,12288,0.0977475576930576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,10240,0.2549022303687202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,12288,0.276180452770657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,65536,1.7230710983276367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,51200,1.249121771918403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,10240,0.08238044712278578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,8192,0.20594844553205702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,12288,0.30835021866692436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,8192,0.07511288589901395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,10240,0.24853955374823677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,7168,0.1751404470867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,8192,0.18484177854326037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,10240,0.25222576989067924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,7168,0.06823199987411499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,8192,0.2064453363418579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,5120,0.11881066693200006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,7168,0.22108888626098633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,6144,0.05440088775422838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,7168,0.170960889922248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,6144,0.14368266529507107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,4096,0.09800355301962958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,5120,0.050055109792285495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,6144,0.14846044116550022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,5120,0.12128178278605144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,3584,0.07776977618535359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,4096,0.038691557115978666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,4096,0.09668533007303874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,5120,0.12832889291975233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,3072,0.06506044334835477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,3584,0.03438577718204922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,4096,0.10682044426600139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,2560,0.056109335687425405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,3584,0.08528711398442586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,3072,0.031810667779710554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,3584,0.09483999676174587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,2048,0.04686577783690559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,2560,0.027853333287768896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,3072,0.08387911319732666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,2560,0.06552533308664958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,1536,0.03398222062322829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,2560,0.07205955849753486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,2048,0.024118221468395654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,6144,0.14200711250305176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,1024,0.025566221939192876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,2048,0.060630222161610924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,1536,0.020067556036843192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,1536,0.04361777835422092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,768,0.02000266644689772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,1536,0.05002133382691277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,1024,0.017910222212473553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,1024,0.032808889945348106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,512,0.015492444237073263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,1024,0.0391813317934672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,768,0.016176889340082806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,768,0.03289688958062066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,512,0.021936888496081035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,256,0.01068800025516086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,512,0.027727110518349543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,3072,0.07504000266393025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,256,0.013773333695199756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,128,0.009559999737474654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,256,0.01905066602759891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,256,0.021922666165563796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,64,0.008277333445019191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,128,0.013800000150998434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,2048,0.05371555685997009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,5120,32,0.008392888638708327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,128,0.01819644371668498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,5120,128,0.01754577789041731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,64,0.018227555685573153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,32,0.018220444520314533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,5120,768,0.027273777458402846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,5120,512,0.014816888504558139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,65536,0.45246134863959414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,65536,1.0231697294447157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,51200,0.7651582294040256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,51200,0.35810221566094297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,16384,0.10585866371790568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,16384,0.34386221567789715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,65536,1.6400506761338975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,16384,0.3197191026475695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,51200,1.2155733108520508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,65536,1.2587902281019423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,51200,1.0042471355862088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,12288,0.08259911007351346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,10240,0.17288355032602945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,12288,0.233696010377672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,8192,0.13991288344065347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,10240,0.07036177979575263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,12288,0.24688977665371367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,10240,0.23524710867140028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,16384,0.2769289016723633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,8192,0.056286222404903836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,10240,0.2024782233768039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,6144,0.10053422053654988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,7168,0.05041422115431892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,7168,0.14314488569895426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,7168,0.14060000578562418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,12288,0.2218479977713691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,6144,0.044043554200066455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,6144,0.1198542250527276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,5120,0.08919199970033433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,6144,0.12318221728006999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,4096,0.06997333632575141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,5120,0.03790488839149475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,5120,0.10007911258273655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,4096,0.03193777799606323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,4096,0.08080977863735624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,4096,0.08662488725450303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,7168,0.12859555085500082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,8192,0.15629066361321345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,3584,0.028844445943832397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,3584,0.07193511062198214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,3584,0.0779306689898173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,8192,0.17346755663553873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,2560,0.044616887966791786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,3072,0.06338577800326876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,3072,0.07031999693976508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,2560,0.05476977758937412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,2560,0.023546665906906128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,2560,0.060489780373043485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,2048,0.02038666605949402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,2048,0.04536799920929802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,1536,0.02773777809407976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,3584,0.05741155809826321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,1536,0.01772177716096242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,1536,0.036706666151682533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,3072,0.0536080002784729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,3072,0.02612888813018799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,1024,0.027608889672491286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,768,0.015139554937680563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,1024,0.015133332875039844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,2048,0.03447466757562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,768,0.02323644525474972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,512,0.011385777758227454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,768,0.014120888378885059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,768,0.02831911047299703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,5120,0.10521333085166083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,2048,0.05241866575347053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,256,0.008647110727098253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,512,0.018938667244381376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,512,0.013008889224794177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,512,0.023597333166334365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,1024,0.018743111027611625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,256,0.012068444656001197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,256,0.016364443633291457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,1536,0.042584889464908175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,64,0.0063164445261160536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,256,0.02085777786042955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,4096,128,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,32,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,128,0.016418667303191293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,64,0.01570044457912445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,4096,1024,0.0332275562816196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,32,0.015538665983412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,65536,0.4221573405795627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,65536,0.9025502734714084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,4096,128,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,51200,0.7220346662733289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,4096,128,0.01550222271018558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,16384,0.2697715494367811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,51200,0.32707643508911133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,65536,1.4080293443467882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,16384,0.09764622317420112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,65536,1.1710969077216256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,16384,0.27792978286743164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,12288,0.20302132765452066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,16384,0.27969243791368276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,12288,0.07689066727956136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,10240,0.17831733491685656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,51200,1.115704854329427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,51200,0.8649777836269803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,12288,0.21724532710181343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,10240,0.08383644289440578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,8192,0.13838044802347818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,10240,0.17613777849409315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,7168,0.1193502214219835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,10240,0.17309066984388563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,8192,0.15990132755703396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,6144,0.10281333658430312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,7168,0.051910221576690674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,7168,0.13957599798838297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,8192,0.14404000176323786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,7168,0.12768977218204075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,5120,0.08398133516311646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,6144,0.04176000091764662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,6144,0.11150577333238389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,4096,0.06888177659776476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,5120,0.03401066528426276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,6144,0.10993866788016425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,5120,0.08996622429953681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,3584,0.05719288852479723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,5120,0.09514666928185357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,4096,0.030010667112138536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,4096,0.07322221994400024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,3072,0.0510044429037306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,12288,0.22326665454440645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,4096,0.07810933060116239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,3584,0.02679288884003957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,3584,0.06504799922307332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,2560,0.03987911012437608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,3584,0.06984977589713202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,2048,0.03279377685652839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,3072,0.05706666575537788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,2560,0.0218231115076277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,3072,0.06300710969501071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,8192,0.05408177773157755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,2560,0.049485332436031766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,1536,0.024960888756646052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,2560,0.055564443270365395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,2048,0.0185244447655148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,1024,0.017092444830470614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,2048,0.04102222124735514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,1536,0.016519111063745286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,768,0.013772444592581855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,2048,0.046494222349590726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,1536,0.033018665181265935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,1536,0.03886577818128798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,1024,0.02460177739461263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,768,0.013109332985348172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,1024,0.030894223186704848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,768,0.021439111895031396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,256,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,768,0.025775111383861963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,512,0.01237422227859497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,512,0.017198221551047433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,128,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,512,0.02181511123975118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,256,0.011908444265524546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,256,0.015004444453451367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,64,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,3072,0.02370044423474206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,128,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,128,0.014397333065668741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,128,0.015453333655993143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,64,0.014138667119873894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3584,32,0.014442667365074158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3584,1024,0.014134221606784396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,512,0.011042666931947073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,65536,0.3639697763654921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,65536,0.891275511847602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,51200,0.5684426625569662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3584,256,0.019351111518012155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3584,32,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,16384,0.21670044793023002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,65536,1.1288586722479927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,51200,0.3742826779683431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,51200,0.8640613555908203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,12288,0.1639066669676039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,16384,0.24557778570387098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,65536,1.1199252870347765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,16384,0.2677084339989556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,10240,0.1413982179429796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,12288,0.06851110855738322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,12288,0.23255109786987305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,51200,0.8746373388502332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,8192,0.11533333195580377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,10240,0.059679110844930015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,12288,0.21560711330837676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,10240,0.1544240050845676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,7168,0.10274133417341445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,8192,0.04761244522200691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,10240,0.18504088454776338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,8192,0.1307013299730089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,6144,0.08675022257698907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,7168,0.04205066627926297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,7168,0.10941244496239556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,6144,0.03661777906947666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,7168,0.1254426638285319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,6144,0.09981955422295465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,16384,0.09307199716567993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,6144,0.10880355702506171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,4096,0.057132442792256675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,5120,0.03223289052645365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,5120,0.08185155524147882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,5120,0.09195288684633042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,4096,0.027274666561020747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,4096,0.0652551121181912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,4096,0.07675200038486056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,3584,0.024889777104059856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,3584,0.0572684473461575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,3584,0.06946666373146905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,8192,0.14383555783165827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,3072,0.05007822314898173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,3072,0.022292445103327434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,3072,0.060876442326439746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,5120,0.05910933017730713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,2048,0.02590311070283254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,2560,0.020249777370029025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,2560,0.04330577784114414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,2560,0.05386755532688565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,2048,0.017683555682500202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,2048,0.036794665786955096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,3584,0.04577689038382637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,2048,0.04484000139766269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,1024,0.0151591118839052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,1536,0.015385778413878547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,1536,0.0295688874191708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,3072,0.03800977932082282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,768,0.012065778175989786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,1536,0.037412444750467934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,1024,0.013424889081054263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,1024,0.02237422267595927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,512,0.009256000320116678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,768,0.012716444830099741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,1024,0.028980443874994915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,768,0.018974221414989896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,256,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,2560,0.032087110810809664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,768,0.025175111161337957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,512,0.01204177737236023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,128,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,512,0.015649777319696214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,512,0.0210817770825492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,64,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,256,0.013356444736321768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,128,0.011221333510345884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,256,0.017994667092959087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,128,0.013075555364290873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,3072,128,0.013600000076823764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,1536,0.020319110817379422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,64,0.01275466630856196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,3072,32,0.01312888910373052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,65536,0.5588497585720485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,65536,0.30850309795803493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,51200,0.46801069047715926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,16384,0.17579910490247938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,51200,0.24527290132310656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,3072,256,0.011062222222487131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,65536,0.9463716083102756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,51200,0.7383031315273709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,3072,32,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,16384,0.07738311423195733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,12288,0.1383839978112115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,65536,0.9357306162516276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,16384,0.2249120076497396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,10240,0.10785866445965236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,16384,0.23616533809238008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,12288,0.1620746718512641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,51200,0.7535217603047689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,8192,0.09441688987943862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,10240,0.056483553515540234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,10240,0.14147199524773493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,7168,0.08044266700744629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,8192,0.040579554107454084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,8192,0.10888266563415527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,8192,0.13665510548485651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,7168,0.03682933251063029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,7168,0.09857422113418579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,7168,0.11151377360026042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,6144,0.032519999477598406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,6144,0.09591822491751777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,5120,0.056292441156175405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,12288,0.060172445244259305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,5120,0.07067288955052693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,12288,0.17797066105736625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,5120,0.08215910858578153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,5120,0.027983999914593164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,10240,0.15281511677636042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,3584,0.039930668142106794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,4096,0.02385066615210639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,4096,0.05707999732759264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,6144,0.07474844323264228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,4096,0.0684915582338969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,3072,0.03414400087462531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,3584,0.05124444431728787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,3584,0.021937777598698933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,2560,0.02959377898110284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,3584,0.060805333985222705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,3072,0.020078221956888836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,6144,0.08675111002392238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,3072,0.04538311229811775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,2048,0.0243857784403695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,3072,0.05465244584613376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,1536,0.01942311061753167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,2560,0.03917422228389316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,2048,0.016207110550668504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,2560,0.04720711045795017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,1024,0.013992889059914483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,2048,0.04052711195415921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,1536,0.014484445254007975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,1536,0.02683111031850179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,4096,0.04511022236612108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,768,0.011328889264000786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,1536,0.033575110965304904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,1024,0.012453333371215396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,1024,0.020030222005314298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,1024,0.02569866677125295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,768,0.011764444410800934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,768,0.016982222596804302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,256,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,768,0.022621333599090576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,512,0.014115555418862237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,512,0.011011555790901184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,128,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,512,0.01924800044960446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,256,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,256,0.012366222010718452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,64,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,256,0.016876444220542908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,32,0.005469333380460739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,128,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,128,0.011695110963450538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2560,2560,0.018304889400800068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2560,128,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,64,0.011583111352390714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,2048,0.03346844514211019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2560,32,0.011752888560295105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2560,512,0.008989333278603023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,65536,0.5617973539564345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,65536,0.2803395589192708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,51200,0.4170488781399197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,16384,0.15257244639926487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,65536,0.7619324260287814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,16384,0.09495644436942206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,51200,0.601219548119439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,16384,0.18537067042456734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,12288,0.10283644331826104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,65536,0.7959013515048556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,16384,0.21825688415103486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,51200,0.6119395362006294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,10240,0.09017866849899292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,12288,0.058021333482530385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,8192,0.07266222106085883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,10240,0.05226666728655497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,12288,0.15956444210476345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,12288,0.15491733286115858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,10240,0.12917866971757677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,7168,0.06580533583958943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,8192,0.03840177920129564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,10240,0.13171288702223036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,8192,0.09394311242633396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,6144,0.05646666553285387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,8192,0.10940355724758571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,7168,0.03955466548601786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,5120,0.045576890309651695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,7168,0.0836755567126804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,6144,0.030974222554100886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,7168,0.09499377674526638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,4096,0.037652442852656044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,6144,0.08189866940180461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,5120,0.02681688964366913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,5120,0.06137422058317396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,3584,0.03136533498764038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,51200,0.22438576486375597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,5120,0.07021511263317533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,4096,0.022446221775478784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,3072,0.027038221557935078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,4096,0.0507919987042745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,3584,0.019181332654423185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,4096,0.05872178077697754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,3584,0.0443466669983334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,2560,0.022495110829671223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,3072,0.017295999659432303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,3072,0.03967555695109897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,2048,0.01942222151491377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,3072,0.0466231107711792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,2560,0.01720888912677765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,2560,0.034028444025251604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,1536,0.015146666102939181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,2560,0.040821333726247154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,2048,0.015642666154437594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,2048,0.028594666057162817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,1024,0.010919111470381418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,2048,0.03513244456715054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,1536,0.01294844432009591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,1536,0.023172444767422144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,768,0.008928889201747047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,6144,0.07322400146060519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,1536,0.028765334023369685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,1024,0.011369778050316704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,1024,0.017672888106769986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,512,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,768,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,1024,0.022682666778564453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,768,0.015223110715548197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,768,0.020278222031063504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,512,0.01239999963177575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,512,0.0184871107339859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,3584,0.05247466762860616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,128,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,256,0.009824000298976898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,256,0.010938666760921478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,64,0.00424177779091729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,256,0.015574221809705099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,128,0.009755555954244401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,32,0.004573333180612988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,128,0.01036000003417333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,2048,128,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,64,0.010279999838935005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,2048,32,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,65536,0.41155023045010036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,65536,0.2337822119394938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,51200,0.29326756795247394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,2048,256,0.005331555588377847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,16384,0.10673066642549302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,2048,512,0.010175111393133799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,65536,0.6109182039896647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,51200,0.18582844734191895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,12288,0.08079644706514147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,16384,0.061864886018964976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,51200,0.49837954839070636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,16384,0.15043200386895075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,65536,0.7438071038987902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,10240,0.07496711280610827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,12288,0.04883111185497708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,16384,0.19502577516767713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,12288,0.11866845024956597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,8192,0.06293511390686035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,51200,0.5834257867601182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,10240,0.04301955633693271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,12288,0.1493591070175171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,10240,0.09607911109924316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,7168,0.05587466557820638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,8192,0.032167110178205706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,6144,0.04696177774005466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,8192,0.07726933558781941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,7168,0.029172443681293066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,8192,0.10423555639055039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,5120,0.03737866547372606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,6144,0.025733333494928148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,7168,0.09072622325685288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,6144,0.05985866652594673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,4096,0.029300444655948218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,6144,0.0801324446996053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,5120,0.0225564440091451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,3584,0.022272888157102797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,5120,0.05065155691570706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,5120,0.06815822256935967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,4096,0.01937599976857503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,3072,0.020042666130595736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,4096,0.05715644359588623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,3584,0.018124444617165458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,3584,0.03762844536039565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,2560,0.017447110679414537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,3584,0.05127822359402975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,3072,0.016711999972661335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,3072,0.033166223102145724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,2048,0.014845333165592618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,2560,0.015154666370815702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,2560,0.028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,10240,0.12755821810828313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,2560,0.03973066806793213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,1536,0.01142133358452055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,2048,0.024358222881952923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,2048,0.014091556270917257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,7168,0.06883733140097724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,1024,0.008467555873923832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,2048,0.033490667740503945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,1536,0.012440000143316058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,1536,0.020008888509538438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,768,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,1536,0.02779910961786906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,1024,0.015312888556056552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,1024,0.02217955556180742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,4096,0.04197688897450765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,768,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,512,0.006501333167155583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,768,0.013079110946920184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,256,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,768,0.019506666395399306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,512,0.00979022185007731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,512,0.011056889262464313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,512,0.01685599982738495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,256,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,256,0.009319110876984065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,3072,0.0455040004518297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,64,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,256,0.01389244364367591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,32,0.004004444513056013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,128,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,128,0.008984000318580205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1536,128,0.011349333657158745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,64,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1536,32,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,65536,0.27418043878343373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1536,1024,0.01107911103301578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,51200,0.22922134399414062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,65536,0.18696355819702148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,16384,0.08063555426067777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,65536,0.4635457992553711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,51200,0.14796622594197592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,51200,0.4013768831888835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,12288,0.05947911077075534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,16384,0.05137066708670723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,16384,0.12028977606031631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1536,128,0.00425511101881663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,65536,0.6029031011793349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,10240,0.04741244514783224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,16384,0.16272444195217556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,12288,0.041166222757763334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,51200,0.4723804261949327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,10240,0.034959110948774554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,12288,0.12322933144039577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,10240,0.07676444451014201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,7168,0.032454222440719604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,8192,0.026299556096394856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,10240,0.10524444447623359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,8192,0.061788446373409696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,6144,0.027098667290475633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,8192,0.08512266476949056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,7168,0.054810666375690036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,5120,0.022887110710144043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,7168,0.07543200254440308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,6144,0.020909332566791113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,6144,0.04734311170048184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,4096,0.018951111369662814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,6144,0.06623378064897326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,5120,0.01864266726705763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,5120,0.04047200083732605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,3584,0.01685688893000285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,4096,0.016488000750541687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,5120,0.05754044320848253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,4096,0.0336151123046875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,3072,0.01514933341079288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,4096,0.048287110196219556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,3584,0.015478221906556023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,8192,0.03762044509251913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,3584,0.030176000462638006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,12288,0.09472355577680801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,2560,0.012836444709036084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,3584,0.042769776450263135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,3072,0.014160888062583076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,3072,0.02664711078008016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,2048,0.010760000182522668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,2560,0.01295733369059033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,3072,0.037885334756639265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,1536,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,2560,0.02337155573897892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,2048,0.011976888610257043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,2560,0.032928887340757586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,7168,0.02371999952528212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,2048,0.02012888921631707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,1536,0.010775999890433418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,1536,0.01664266652531094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,768,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,1024,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,1536,0.024043555061022442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,1024,0.01274400038851632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,1024,0.019501333435376484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,768,0.00958933350112703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,768,0.011323555476135679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,256,0.003847111016511917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,768,0.01756533318095737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,512,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,512,0.015146666102939181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,256,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,128,0.0035253332720862497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,256,0.00814133303032981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,256,0.013100444442696042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,1024,128,0.008615111311276754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,128,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,64,0.003133333391613431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,32,0.003206222214632564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,128,0.01034400032626258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,1024,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,1024,2048,0.028425776296191748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,64,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,65536,0.24704445732964408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,32,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,1024,512,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,51200,0.18433777491251627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,65536,0.18386933538648817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,65536,0.41581421428256565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,16384,0.061463998423682324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,51200,0.16499111387464735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,51200,0.3267039987776014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,12288,0.04575555523236593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,16384,0.053871999184290566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,65536,0.5953840149773492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,16384,0.10848710934321086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,51200,0.46826309627956814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,10240,0.03846133417553372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,16384,0.1596062183380127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,12288,0.040172444449530706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,12288,0.08224799897935656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,8192,0.02900444467862447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,10240,0.03481244378619724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,12288,0.12514489226871067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,10240,0.06860355536142985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,7168,0.025962667332755193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,10240,0.10413777828216553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,8192,0.026014222039116755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,8192,0.05543733305401272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,6144,0.021789333886570398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,8192,0.08385155598322551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,7168,0.02313066687848833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,7168,0.04908355408244663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,5120,0.02004888819323646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,1024,512,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,7168,0.07489689191182454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,6144,0.043059554364946157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,4096,0.016128000285890367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,6144,0.06554666492674086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,5120,0.01847555571132236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,3584,0.014334221680959066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,5120,0.03684622380468581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,4096,0.01649777756796943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,5120,0.05627999703089396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,3072,0.012864888542228274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,4096,0.030525333351559106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,3584,0.015142222245534262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,4096,0.04672000143263075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,2560,0.011578666667143503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,3584,0.02752711044417487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,3072,0.014124444789356656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,3584,0.04177155428462558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,2048,0.009663999908500248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,2560,0.013080000049538083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,3072,0.03759555684195624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,1536,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,2560,0.021527111530303955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,2560,0.03208533260557387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,2048,0.011807999677128263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,2048,0.018380444910791185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,1024,0.0063928887248039246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,1536,0.010846222440401712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,1536,0.015437333120240105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,768,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,1536,0.023369777533743117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,1024,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,6144,0.02115999990039402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,512,0.0046160001721647055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,1024,0.018577777677112155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,768,0.009559111462699043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,768,0.010417777631017896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,256,0.0035013332962989807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,768,0.01686311099264357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,512,0.009008888569143083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,512,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,128,0.0031991110493739447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,512,0.014474666780895658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,256,0.008624888956546783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,256,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,3072,0.024528000089857314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,64,0.0028604444944196274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,256,0.012697777814335294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,768,32,0.0031591111587153543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,768,128,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,128,0.007019555403126611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,128,0.010326222413116032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,65536,0.18436088826921251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,64,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,32,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,768,2048,0.027462222509913977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,51200,0.14749688572353786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,65536,0.13886488808525935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,16384,0.046332445409562856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,768,1024,0.011975111232863532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,51200,0.11157333850860596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,65536,0.337024900648329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,12288,0.03243288728925917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,51200,0.2649244467417399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,16384,0.0391173329618242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,65536,0.45084089703030056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,16384,0.0889822244644165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,10240,0.026006221771240234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,51200,0.35892711745368105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,12288,0.0673546658621894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,16384,0.12439021799299453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,8192,0.020039111375808716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,10240,0.0270444436205758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,10240,0.05675111214319865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,12288,0.0959928896692064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,7168,0.016979555288950603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,8192,0.01941066649225023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,8192,0.04569777846336365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,6144,0.0151573336786694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,8192,0.06642044252819486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,7168,0.058561775419447154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,6144,0.035862220658196345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,5120,0.013080000049538083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,6144,0.051979555024041064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,5120,0.03068444463941786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,5120,0.044622222582499184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,5120,0.014429334137174817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,12288,0.030862222115198772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,4096,0.011076444553004371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,4096,0.012669333153300814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,10240,0.08093688885370891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,3584,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,4096,0.025399110383457605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,4096,0.03734577695528666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,7168,0.04080977704789903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,3072,0.008861333131790161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,3584,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,3584,0.023166222704781428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,2560,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,3584,0.03327111072010464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,6144,0.015909334023793537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,3072,0.020642666353119742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,2048,0.006567111031876669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,3072,0.0295688874191708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,2560,0.010741333166758219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,1536,0.005728000154097875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,2560,0.017906667457686532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,2560,0.026549332671695288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,2048,0.010032888915803697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,2048,0.015599111715952555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,2048,0.022742221752802532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,1536,0.00944266633854972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,1536,0.012999999854299756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,768,0.003906666818592283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,1536,0.019553777244355943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,1024,0.008997333546479544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,1024,0.010392000277837118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,512,0.0035093331502543557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,1024,0.01585244470172458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,768,0.008642666869693333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,768,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,256,0.002871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,7168,0.01759999990463257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,768,0.01478666729397244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,512,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,512,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,128,0.0028435554769304064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,256,0.007674666742483775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,512,0.012896888785892062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,64,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,3072,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,32,0.0027013334135214486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,256,0.011394666300879585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,128,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,512,128,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,65536,0.10897422499126858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,64,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,512,1024,0.00452533322903845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,32,0.006217777729034424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,51200,0.0856319997045729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,65536,0.1380008856455485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,65536,0.2945600085788303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,51200,0.1104924413892958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,51200,0.23092267248365614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,65536,0.4380933443705241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,12288,0.02492977844344245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,51200,0.3447466691335042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,16384,0.03864533371395535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,16384,0.11985333760579427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,10240,0.020152888364262052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,512,256,0.006776000062624614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,12288,0.058339556058247886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,512,128,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,8192,0.01650044487582313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,12288,0.09292799896664089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,10240,0.027404444085227117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,10240,0.049398223559061684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,7168,0.014690667390823364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,10240,0.07819110817379422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,8192,0.019467555814319186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,8192,0.040427555640538536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,8192,0.06360355350706312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,16384,0.031544887357287936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,6144,0.013054221868515015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,7168,0.01756000022093455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,7168,0.03583911061286926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,5120,0.011057777537239922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,7168,0.056256890296936035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,6144,0.03166488806406657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,16384,0.0766400032573276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,4096,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,5120,0.027179555760489568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,5120,0.014341332846217685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,12288,0.030227555168999568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,5120,0.042407999436060585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,3584,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,4096,0.012655110822783576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,4096,0.022442667020691767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,3072,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,4096,0.03514666689766778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,3584,0.012139555480745105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,3584,0.020397333635224234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,2560,0.0058808889653947614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,3584,0.031594667169782854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,3072,0.011412444214026133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,3072,0.0284897784392039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,2048,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,2560,0.010728889041476779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,2560,0.016167110866970487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,2560,0.025397333833906386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,1536,0.004218666503826777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,6144,0.016029333074887592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,2048,0.01034577770365609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,2048,0.01404355631934272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,1024,0.003552000141806073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,2048,0.022097777989175584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,6144,0.05007466673851013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,1536,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,768,0.0032266666077905228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,1536,0.011744000017642975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,1024,0.00869422240389718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,1536,0.01890222231547038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,512,0.0028515555378463534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,1024,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,1024,0.015491555134455362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,768,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,256,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,768,0.014152889450391134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,512,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,128,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,512,0.012436444560686747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,256,0.007667555577225155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,3072,0.018239110708236694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,64,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,256,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,256,32,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,256,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,256,128,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,128,0.00562755556570159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,65536,0.07697244485219319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,256,128,0.00832177781396442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,64,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,32,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,65536,0.13673067092895508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,65536,0.252491553624471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,16384,0.02572533322705163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,51200,0.19795110490587023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,51200,0.10980088843239678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,768,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,65536,0.43366045422024196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,12288,0.017641777793566387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,51200,0.3413377867804633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,16384,0.0662462247742547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,256,512,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,10240,0.01368799971209632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,16384,0.1186186605029636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,12288,0.031031999323103163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,12288,0.05080266793568929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,8192,0.012840888566441007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,12288,0.09159288803736369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,10240,0.04280799958440992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,7168,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,10240,0.07653688722186618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,8192,0.03489066825972663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,8192,0.020272889071040683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,51200,0.06091377470228407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,6144,0.010039110978444418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,8192,0.06297955248090956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,7168,0.01759822169939677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,5120,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,7168,0.055789331595102944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,6144,0.01679733395576477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,6144,0.027454222242037456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,4096,0.007176000210973952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,6144,0.04900000161594815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,5120,0.014878221684032016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,5120,0.023569777607917786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,3584,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,16384,0.038591110044055514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,4096,0.013409777647919126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,5120,0.04179999894566006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,4096,0.01997599999109904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,3072,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,4096,0.034781333472993635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,3584,0.012230222423871359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,3584,0.018216000662909616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,2560,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,10240,0.02604266670015123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,3584,0.03127377894189622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,3072,0.01631644368171692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,3072,0.011424889167149862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,2048,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,3072,0.028507557180192735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,2560,0.011120888921949597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,1536,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,2560,0.01447111037042406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,2048,0.012796444197495779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,2048,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,7168,0.03127377894189622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,1024,0.003528888854715559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,2048,0.021612443857722815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,1536,0.009650666680600908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,1536,0.010782221953074137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,768,0.0031804444475306403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,1536,0.01861422260602315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,1024,0.008987555901209513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,512,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,1024,0.015501333607567681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,768,0.008452444440788692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,768,0.00794222205877304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,256,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,768,0.013771555489963956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,512,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,512,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,128,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,512,0.012152888708644442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,256,0.007700444095664554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,256,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,64,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,256,0.010344888601038191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,128,32,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,128,0.007673777639865875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,128,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,65536,0.06979111168119642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,64,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,51200,0.05642844571007622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,2560,0.02478666603565216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,16384,0.022395556171735127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,12288,0.013487999637921652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,65536,0.252621332804362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3328,128,1024,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,10240,0.011800888511869641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,16384,0.06620799832873873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,8192,0.012509333590666452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,51200,0.1977457735273573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,7168,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,12288,0.05053244365586174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,6144,0.009383999639087254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,5120,0.008303110798199972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,10240,0.04252800014283922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,4096,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,7168,0.03091111116939121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,6144,0.027071111732059058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,3072,0.005745777653323279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,5120,0.023384888966878254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,2560,0.005155555489990446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,4096,0.01980977753798167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,2048,0.004535111288229625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,3584,0.01812266641192966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,1536,0.0038444445365005066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,3072,0.0162426663769616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,1024,0.0031991110493739447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,2560,0.01444355481200748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,768,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,128,32,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,512,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,2048,0.012723555167516073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,256,0.002464888824356927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,1536,0.010759111079904767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,128,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,1024,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,64,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,768,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,32,0.002191999927163124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,512,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,256,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,65536,0.0749431120024787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,128,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,51200,0.05736088752746582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,64,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,16384,0.020482665962643094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,32,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,12288,0.01108000013563368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,64,3584,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,10240,0.010028444230556488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,8192,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,65536,0.2519520123799642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,16384,0.06641244226031832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,7168,0.008794666992293464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,12288,0.05062133404943678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,6144,0.008350222474998897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,8192,0.034576889541414045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,5120,0.010397333237859938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,7168,0.03068800104988946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,4096,0.008748444418112436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,6144,0.027126222848892212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,3584,0.007932444413503012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,5120,0.023370666636361018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,3072,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,4096,0.019896888070636325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,2560,0.006450666735569636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,3584,0.018093332648277283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,2048,0.004538666870858935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,3072,0.01647111111217075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,64,8192,0.03455555438995361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,1536,0.003830222205983268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,1024,0.003156444471743372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,2560,0.014517333772447375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,768,0.0028488888508743713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,2048,0.012746666868527731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,512,0.00249422217408816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,1536,0.01076533314254549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,256,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,1024,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,10240,0.04252800014283922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,128,0.002157333410448498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,768,0.007684444387753804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,64,0.0019697778754764134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3328,32,32,0.0018648888087934917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,512,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,256,0.005633777628342311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,128,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,64,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,32,0.005000000198682149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3328,32,51200,0.19767999649047852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,16384,1.3265359666612413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,12288,2.8295413123236766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,16384,3.7385982937282987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,16384,4.3703837924533415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,12288,0.9599972830878364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,51200,5.021412319607205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,12288,3.2519796159532337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,10240,0.7959991031222873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,10240,2.8572425842285156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,10240,2.5421199798583984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,8192,0.6838835610283746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,7168,0.5941973262363011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,7168,1.8677955203586156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,8192,2.003063201904297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,8192,2.438324398464627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,7168,1.695271068149143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,6144,0.5026168823242188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,5120,0.4191768964131673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,6144,1.5661822424994574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,5120,1.4787537256876628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,6144,1.4797368579440648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,5120,1.198764483133952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,4096,0.372648901409573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,4096,1.1580097410413954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,3584,0.8177404403686523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,4096,0.9889520009358724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,3584,0.3144320117102729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,3584,0.8332515292697482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,3072,0.281422217686971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,3072,0.8737422095404731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,51200,12.3387451171875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,2560,0.2436444494459364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,3072,0.7645564609103732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,2560,0.6128684679667155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,2048,0.20681956079271105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,2048,0.5058631367153591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,1536,0.35093066427442765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,1536,0.16652444998423258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,1024,0.2614906628926595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,2048,0.5126498010423448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,768,0.19458311133914521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,1024,0.1309937768512302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,1024,0.32859910859002006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,512,0.15554577774471706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,2560,0.647955576578776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,768,0.1158755620320638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,256,0.10936977465947469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,768,0.2389235496520996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,512,0.1050213310453627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,512,0.20416711436377633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,128,0.08426311280992295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,256,0.09133066733678182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,64,0.06984444459279378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,256,0.18404622872670492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,65536,32,0.07278044356240167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,128,0.17319733566708037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,64,0.17621955606672499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,1536,0.39600088861253524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,32,0.17927555243174234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,65536,128,0.0898924469947815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3328,128,128,0.008245333201355403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,65536,51200,20.029400295681423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,51200,3.8481788635253906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,65536,5.162025027804904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,16384,3.1462186177571616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,16384,0.9565618303087023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,12288,2.2274292839898004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,16384,3.4743245442708335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,51200,9.519557529025608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,12288,0.7463235325283475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,12288,2.6450222863091364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,10240,1.868554645114475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,10240,0.6162106725904677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,8192,0.5066702100965712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,10240,2.019531461927626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,8192,1.8180792066786025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,8192,1.6513377295600042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,7168,0.45603201124403214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,65536,12.390268113878038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,7168,1.5688435236612956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,6144,0.39247645272148984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,6144,1.1716204749213326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,7168,1.3918737835354273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,5120,1.0190756056043837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,5120,0.3353448973761664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,6144,1.1605182223849826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,4096,0.8117021984524198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,5120,1.007812394036187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,4096,0.7672959963480631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,3584,0.23976445198059082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,3584,0.7734337912665473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,3584,0.6648711098564996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,3072,0.6445537673102485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,3072,0.21674844953748915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,2560,0.5440337922837999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,2560,0.20775111516316733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,3072,0.5778089099460179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,2560,0.4924000104268392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,2048,0.46799908743964297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,51200,14.919508192274305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,2048,0.17127111223008898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,1536,0.2956284417046441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,4096,0.2834471066792806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,1024,0.2562648985120985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,2048,0.47470935185750324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,1536,0.3116942246754964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,768,0.1944319936964247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,1024,0.2429342269897461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,512,0.1265208853615655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,768,0.09216977490319146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,256,0.08140444755554199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,512,0.08506221903695001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,768,0.21969600518544516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,128,0.06359377834531996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,512,0.16086578369140625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,256,0.0736408895916409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,64,0.05521066652403938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,51200,32,0.05774844355053372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,128,0.0721199976073371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,128,0.14045333862304688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,256,0.14398933781517878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,64,0.13886133829752603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,32,0.14079022407531738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,1536,0.15862933794657388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,51200,1024,0.10380533006456162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,51200,65536,17.899356418185764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,51200,1.23143736521403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,65536,1.6607173283894856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,16384,0.39087555143568253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,16384,0.9174408382839627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,16384,1.116728040907118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,51200,2.9860827128092446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,12288,0.2554302215576172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,12288,0.8812453481886121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,10240,0.21600266297658285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,10240,0.6696071094936795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,8192,0.5760515530904134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,65536,3.9771376715766062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,8192,0.2009591129091051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,10240,0.677296002705892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,8192,0.5293164253234863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,7168,0.18063199520111084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,7168,0.5417813195122613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,6144,0.36262043317159015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,7168,0.4569279882642958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,6144,0.15938133663601345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,12288,0.8289955457051595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,5120,0.4204382101694743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,6144,0.38314223289489746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,5120,0.11554488870832656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,4096,0.2823066711425781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,51200,4.800743103027344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,5120,0.3285875585344103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,4096,0.16442399554782444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,3584,0.24118844668070474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,3584,0.0866862204339769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,3072,0.21947288513183594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,4096,0.25158222516377765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,2560,0.17394755946265328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,3584,0.22157511446211073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,3072,0.14041599962446424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,2048,0.1288693348566691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,2560,0.08283022377226087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,3072,0.1957475609249539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,2048,0.05760622024536133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,2560,0.17968355284796822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,1024,0.07288710938559638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,1536,0.05939110782411364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,2048,0.1583955552842882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,1536,0.1073839995596144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,768,0.06813689072926839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,512,0.0532915559079912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,65536,6.161532508002387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,1024,0.08425066868464152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,256,0.03075466553370158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,768,0.06662578052944608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,512,0.03786311215824551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,128,0.026019554999139573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,512,0.05911733044518364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,256,0.03013333347108629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,64,0.020043555233213637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,256,0.051056000921461314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,128,0.029520889123280842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,32,0.022656889425383672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,128,0.04935555656750997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,64,0.04889600144492256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,16384,32,0.04930577675501505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,16384,1536,0.13768443796369764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,1024,0.03930044505331252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,16384,768,0.036551998721228704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,16384,0.794091542561849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,65536,1.1961386998494465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,51200,0.8785662121242948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,16384,0.29245244132147896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,16384,0.8268586794535319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,12288,0.19958221912384033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,51200,2.482309341430664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,65536,2.8551928202311196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,10240,0.4555653466118707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,10240,0.18684444162580702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,10240,0.4965155389573839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,8192,0.4523715443081326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,8192,0.13355555799272326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,8192,0.4075617790222168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,7168,0.3765564494662815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,7168,0.12008710702260335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,12288,0.6338337792290581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,6144,0.2730240027109782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,6144,0.10520711210038926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,7168,0.38711200820075137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,5120,0.2599991162618001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,12288,0.6530639860365126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,51200,3.9704852634006076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,4096,0.1960000064637926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,6144,0.40079911549886066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,4096,0.07408444748984443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,3584,0.1899475521511502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,5120,0.25427378548516166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,4096,0.1961724493238661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,3584,0.08248533142937554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,3072,0.14686311615837946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,65536,4.473681979709201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,3072,0.059149331516689725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,3584,0.17807377709282768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,2560,0.05249866512086657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,2048,0.10932711097929214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,3072,0.17809955279032388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,1536,0.07841600312127008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,2048,0.04661688870853848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,2560,0.12818488809797499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,1024,0.05143910977575514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,1536,0.041893333196640015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,2048,0.10506400134828354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,1024,0.03183644347720676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,1536,0.08808177709579468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,512,0.033969776497946844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,1024,0.0654639999071757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,768,0.029312888781229656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,256,0.023144889209005568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,768,0.052130666044023305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,128,0.01792888840039571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,512,0.043775998883777194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,256,0.024321777953041926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,64,0.016172443826993305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,256,0.03908888830078973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,128,0.023357333408461675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,32,0.01639111174477471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,128,0.03720266620318095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,2560,0.12159910466935898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,64,0.037419554260041975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,12288,32,0.037662221325768366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,12288,768,0.04319555560747782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,512,0.02773155603143904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,12288,5120,0.09075022406048244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,65536,0.9977430767483182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,51200,0.7598213089836968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,16384,0.6611048910352919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,16384,0.681146674686008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,16384,0.28251555230882436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,51200,1.9085990058051214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,12288,0.5967466566297743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,12288,0.16906489266289604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,65536,2.4164106580946183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,12288,0.6046755578782823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,10240,0.5071004231770834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,8192,0.3760533332824707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,10240,0.42525066269768613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,8192,0.1375759972466363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,8192,0.3551164468129476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,7168,0.3454186651441786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,7168,0.10991022321912979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,6144,0.24054932594299316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,7168,0.28715022404988605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,6144,0.09681244691212972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,5120,0.19068355030483672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,6144,0.34792976909213597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,5120,0.07634933127297296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,4096,0.15115733941396078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,5120,0.20502400398254395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,10240,0.17191467020246717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,4096,0.06365511152479383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,3584,0.1302808920542399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,4096,0.17622133096059164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,3584,0.1446453332901001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,65536,4.368426852756076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,3072,0.05128355489836799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,3072,0.13173511293199328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,2560,0.10359022352430557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,51200,2.9220453898111978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,2048,0.07747199800279406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,2560,0.04544088906712002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,1536,0.0610862241850959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,2048,0.039854221873813205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,2560,0.1197626723183526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,1024,0.04504977663358053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,2048,0.08943466345469157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,1536,0.03368355499373542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,768,0.03561688794030084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,1024,0.026732444763183594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,768,0.02409155501259698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,1024,0.05407199925846524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,512,0.026360889275868733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,256,0.018787556224399142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,768,0.04449333416091072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,128,0.01594311164485084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,512,0.02165422174665663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,3072,0.12402133146921794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,256,0.01980266637272305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,64,0.01333333303531011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,10240,32,0.013354666531085968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,128,0.01980799933274587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,256,0.031599998474121094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,128,0.030563556485705908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,64,0.02994489007525974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,32,0.030340443054835003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,1536,0.0711759991115994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,10240,3584,0.056741330358717174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,65536,0.7764790852864584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,10240,512,0.035763555102878146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,51200,0.6310693422953287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,16384,0.18074133661058214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,16384,0.5196711222330729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,51200,1.703147570292155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,65536,1.9513840145534938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,12288,0.14123911327785915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,12288,0.4565413263108995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,16384,0.6244284311930338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,12288,0.3979395495520697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,10240,0.3564799891577826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,10240,0.11550222502814399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,8192,0.3090764416588677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,8192,0.09461866484747992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,10240,0.3625626564025879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,8192,0.2758693430158827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,7168,0.10195200310813056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,6144,0.20991199546390107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,51200,2.3827707502577042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,7168,0.2361831135219998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,6144,0.07355378071467082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,5120,0.2038737800386217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,5120,0.06683378087149726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,4096,0.12995644410451254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,65536,2.9753074645996094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,6144,0.20632978280385336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,3584,0.11622399753994411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,4096,0.06281510988871257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,3072,0.0973937776353624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,3584,0.05008177624808418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,2560,0.08774399757385254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,4096,0.13619555367363825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,3072,0.05193155672815111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,3584,0.11937067243787979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,2048,0.06999822457631429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,3072,0.10537955496046279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,1536,0.049307554960250854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,2560,0.09396977557076348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,1024,0.03441155619091458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,2048,0.0766871107949151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,1536,0.027224888404210407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,768,0.02801777919133504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,7168,0.24965778986612955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,1536,0.05923733446333143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,1024,0.022456000248591106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,512,0.020940444535679285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,1024,0.045159111420313515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,768,0.020564445190959506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,256,0.015602666470739575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,768,0.037243554989496865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,512,0.018457776970333524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,512,0.029902223083708022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,128,0.01148444414138794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,256,0.017197334104114108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,5120,0.1855315499835544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,256,0.02583733366595374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,64,0.010480889015727572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,8192,32,0.01036800030204985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,128,0.02500711050298479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,64,0.025062221619817946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,2560,0.037512001064088606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,8192,32,0.02472711106141408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,2048,0.03246577911906772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,51200,0.533832867940267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,65536,0.6660408973693848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,8192,128,0.017172444197866652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,16384,0.15586755010816786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,16384,0.5050942103068033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,51200,1.7176906797620985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,65536,2.0176703135172525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,16384,0.5526720152960883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,12288,0.13679822285970053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,12288,0.35955556233723956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,10240,0.10278666681713527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,12288,0.3580079873402913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,8192,0.27485156059265137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,10240,0.3428364329867893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,7168,0.18949688805474174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,8192,0.2575146622127957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,10240,0.3807164563073053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,6144,0.1578577756881714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,65536,2.7650445302327475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,6144,0.06466133064693876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,5120,0.13139377699957952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,6144,0.183098660575019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,4096,0.11007377836439346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,5120,0.05605866511662801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,3584,0.10525333219104344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,5120,0.1757377783457438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,4096,0.04725066820780436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,4096,0.12160533004336888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,3584,0.04237511091762119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,51200,1.9396071963840062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,3584,0.10665066374672784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,3072,0.038274665673573814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,2560,0.07313511106703016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,3072,0.09291199843088786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,8192,0.08312444554434882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,7168,0.09003911415735881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,1536,0.04231377773814731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,2560,0.07983111010657416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,2048,0.028872887293497723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,7168,0.2446213298373752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,1024,0.031130668189790513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,2048,0.06651822063657972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,1536,0.0248346659872267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,768,0.023689778314696416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,1536,0.052673776944478355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,512,0.017685333887736004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,768,0.018533334136009216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,1024,0.04048000110520257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,256,0.012936000194814471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,768,0.033369776275422834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,512,0.01687466601530711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,128,0.01128000020980835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,512,0.02664088871743944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,3072,0.08373066451814438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,256,0.01571999986966451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,64,0.00958577791849772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,256,0.023358222511079576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,32,0.011384888655609555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,128,0.015561777684423657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,128,0.022456000248591106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,64,0.022502221994929846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,7168,2048,0.05455466773774889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,2560,0.03344355689154731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,7168,32,0.02235288918018341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,7168,1024,0.020228443874253165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,16384,0.386136876212226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,51200,0.5460844569736057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,16384,0.15008889304267034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,65536,0.6836320029364692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,16384,0.41722843382093644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,12288,0.13035555680592856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,65536,1.498398251003689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,12288,0.3215973377227783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,51200,1.3989689085218642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,12288,0.32279909981621635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,10240,0.20834755897521973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,8192,0.21715288692050508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,10240,0.10530577765570746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,7168,0.15387733777364096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,8192,0.07532622416814168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,10240,0.2813813421461317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,51200,1.599122683207194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,7168,0.07446222172843085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,8192,0.26380711131625706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,6144,0.059643553362952344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,5120,0.14058400524987116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,6144,0.15606133143107095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,4096,0.11346844832102458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,65536,2.590066697862413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,5120,0.13850577672322592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,4096,0.04495733314090305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,3584,0.10973244243197972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,4096,0.10571910937627156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,3584,0.03978577918476529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,3584,0.0937439997990926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,3072,0.03513866662979126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,3072,0.08175644609663221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,2560,0.030625777112113103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,6144,0.14860533343421087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,2560,0.07001600000593397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,2048,0.05681600173314413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,1536,0.037179556157853864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,2048,0.025979555315441553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,5120,0.05141511228349474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,2048,0.0580622222688463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,1024,0.028802666399214003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,1536,0.022096888886557683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,768,0.020428443948427837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,1536,0.04713955521583557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,1024,0.018390221728218924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,512,0.016389333539538913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,768,0.01682577696111467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,1024,0.035255998373031616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,3072,0.08929066525565253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,256,0.014817777607176038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,512,0.01564088960488637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,768,0.030291554000642564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,128,0.009375111096435124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,64,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,512,0.02369511127471924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,32,0.008599110775523717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,256,0.014478221535682678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,256,0.020748444729381137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,6144,2560,0.068977779812283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,6144,128,0.014464888307783337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,128,0.019938665959570143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,64,0.019951999187469482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,32,0.019870221614837646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,6144,7168,0.25307999716864693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,16384,0.3283289008670383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,65536,0.5394781960381402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,51200,0.43686578008863663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,16384,0.12705600261688232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,16384,0.3542799949645996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,12288,0.3118506537543403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,12288,0.09633511304855347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,10240,0.20344355371263292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,65536,1.384892463684082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,12288,0.32056978013780385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,10240,0.08216710885365804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,8192,0.1701351139280531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,8192,0.06533866458468966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,10240,0.30332620938618976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,51200,1.569127082824707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,8192,0.25444621509975857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,6144,0.12430044015248616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,7168,0.17110933197869194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,65536,1.866895039876302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,5120,0.10790399710337321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,6144,0.06989066468344794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,4096,0.08296710915035672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,6144,0.14927111731635198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,5120,0.04818311002519396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,51200,0.8423075675964355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,3584,0.07272799809773763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,4096,0.042216890388064914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,3072,0.05912711222966512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,4096,0.10842044485939874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,2560,0.050444444020589195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,3584,0.08451200193829006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,3072,0.031091554297341242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,2048,0.04212888744142321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,3072,0.07217244307200114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,2560,0.027775110469924078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,1536,0.035214222139782377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,2048,0.023710222707854375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,7168,0.15386311213175455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,1024,0.02466222147146861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,1536,0.020153777466879953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,7168,0.0591742197672526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,768,0.019911110401153564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,1024,0.017666666044129264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,1024,0.0306968887646993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,512,0.014966222974989148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,768,0.015954666667514376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,768,0.025637333591779072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,256,0.01032977799574534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,5120,0.11759111616346572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,512,0.014472888575659858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,128,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,256,0.01347644461525811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,256,0.018246221873495314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,64,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,128,0.013632888595263163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,128,0.017475555340449016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,5120,32,0.007962666451931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,2560,0.06109688679377238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,2048,0.051289776961008705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,64,0.017489777670966256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,32,0.017235555582576327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,1536,0.04120088948143853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,65536,0.45371468861897785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,5120,3584,0.03398399882846408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,16384,0.22891823450724283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,65536,1.1363804075453017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,51200,0.3537608782450358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,16384,0.13589955700768366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,12288,0.20566488636864555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,51200,1.147396405537923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,16384,0.2970657878451877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,65536,1.5837760501437719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,10240,0.16414933734469944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,12288,0.09165066480636597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,8192,0.14257866806454128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,12288,0.29071201218499076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,10240,0.07953955729802449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,5120,512,0.020848888489935134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,10240,0.18276711304982504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,7168,0.12834221786922878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,8192,0.06269777483410306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,8192,0.15474843978881836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,6144,0.09415377510918511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,7168,0.048567109637790255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,5120,0.08590488963656956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,51200,0.8674231105380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,7168,0.14300178156958684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,4096,0.06584711207283868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,3584,0.05928888585832384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,6144,0.11570310592651367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,4096,0.03131466772821214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,3072,0.04884533418549431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,3584,0.0291244453854031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,4096,0.07653866873847114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,3584,0.06863466898600261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,2560,0.04086044430732727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,3072,0.025504888759719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,2048,0.03262133399645487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,3072,0.06006044149398804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,2560,0.02330488959948222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,1536,0.02699288891421424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,2048,0.02023555503951179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,2560,0.05183466606669956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,1024,0.018968888454967074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,2048,0.04330577784114414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,768,0.015022221538755627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,1536,0.017677333619859483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,1024,0.01516533394654592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,512,0.011197333534558615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,768,0.01408622165520986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,768,0.022287999590237934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,256,0.008583111067612966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,512,0.01315111087428199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,6144,0.04385244515207079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,512,0.01798488861984677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,128,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,5120,0.03645066751374139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,256,0.012102222277058495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,64,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,5120,0.09463022152582805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,4096,32,0.006447111152940326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,4096,128,0.012143111063374413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,256,0.015537778536478678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,128,0.014871999621391296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,64,0.015167999598715039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,32,0.014934221903483072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,65536,0.4094373385111491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,1536,0.03491733471552531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,51200,0.3209555678897434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,65536,0.8221431308322482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,16384,0.25231289863586426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,51200,0.8038889037238227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,12288,0.18654933240678576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,16384,0.09627111090554132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,16384,0.2604195541805691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,10240,0.16724356015523276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,51200,0.9729262457953559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,12288,0.2636328803168403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,10240,0.06447199980417888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,65536,1.2818898095024955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,8192,0.13990133338504368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,10240,0.16427111625671387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,7168,0.12062933709886338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,6144,0.08771999677022298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,8192,0.06614844666586982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,7168,0.05182488759358724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,5120,0.08381155464384292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,7168,0.12109511428409153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,6144,0.039798223310046725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,4096,0.06452622016270955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,6144,0.09992177618874444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,5120,0.03442755672666762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,5120,0.08815199799007839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,3584,0.054337776369518705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,4096,1024,0.02607911162906223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,4096,0.028981334633297388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,3072,0.04601688848601448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,4096,0.06913510958353679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,3584,0.026872888207435608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,2560,0.039156443542904325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,3584,0.061221334669325084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,3072,0.022944889134830896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,2048,0.029299557209014893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,12288,0.08972889184951782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,2560,0.021176889538764954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,1536,0.023970666858885024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,2560,0.047043555312686496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,2048,0.018332444959216647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,1024,0.016701334052615695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,1536,0.01614755557643043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,768,0.013628444737858243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,1536,0.03142311175664266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,1024,0.023619555764728125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,512,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,768,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,8192,0.13422844145033094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,768,0.02029244436158074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,256,0.007968000239796108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,512,0.01236533373594284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,256,0.011696000066068439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,128,0.007336888876226213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,512,0.01646311084429423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,64,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,256,0.014496000276671516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,128,0.011515555282433828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3584,32,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,128,0.013776000175211163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,64,0.013831111292044321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,3072,0.054280890358818897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,32,0.013794666363133324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3584,2048,0.03952711158328586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3584,1024,0.014039110806253223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,65536,0.35806043942769367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,51200,0.5613822407192653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,16384,0.22635467847188315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,51200,0.27885156207614475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,65536,0.9311484230889214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,12288,0.16774933867984346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,16384,0.08836622370613946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,12288,0.08604800038867527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,16384,0.2476417753431532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,10240,0.1154435608122084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,65536,1.015045377943251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,51200,0.8195217980278863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,12288,0.17256711588965523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,8192,0.1185377836227417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,10240,0.0595102243953281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,7168,0.09493155611885919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,8192,0.04539822207556831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,10240,0.1550862259334988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,6144,0.0913662248187595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,5120,0.06562933656904432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,7168,0.04169422056939867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,8192,0.11896977159712051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,6144,0.03794044587347243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,4096,0.04916710986031426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,7168,0.11132621765136719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,3584,0.04329866502020094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,5120,0.03854755560557047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,6144,0.09182311428917779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,4096,0.027112000518374976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,5120,0.0788737800386217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,2560,0.031981332434548274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,3584,0.025913778278562758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,3072,0.022056000100241765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,3584,0.05341688791910807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,2048,0.028312000963422988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,1536,0.021809778279728357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,3072,0.04855288730727302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,2560,0.019888889458444383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,1024,0.01623911162217458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,2048,0.017509332961506315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,2560,0.04122577773200141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,1536,0.015126221709781222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,768,0.013402666482660504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,1024,0.01312888910373052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,1536,0.028786665863460962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,512,0.0101742222905159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,1024,0.021384888225131567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,768,0.012724444270133972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,256,0.0075911109646161394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,768,0.018013333280881245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,512,0.011699555648697747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,128,0.00693866651919153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,512,0.014870221416155497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,64,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,256,0.012886222038004132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,32,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,128,0.011015111373530494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,128,0.012661333713266583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,3072,3072,0.03694577680693732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,64,0.01240355521440506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,4096,0.06201599703894722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,32,0.012712889247470431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,65536,0.29574399524264866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,51200,0.40248444345262313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,3072,2048,0.034464889102511935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,65536,0.6611830923292372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,16384,0.15212621953752306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,16384,0.07863733503553602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,12288,0.11402755313449436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,51200,0.2346115642123752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,3072,256,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,65536,0.8575377994113498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,16384,0.20054311222500273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,10240,0.10261778036753337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,8192,0.07094489203559028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,12288,0.06228444311353895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,10240,0.05239822136031257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,12288,0.19592177867889404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,8192,0.03936266567971971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,8192,0.1064293318324619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,6144,0.05841155846913656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,7168,0.03603822324011061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,5120,0.05230666531456841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,7168,0.09171999825371636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,6144,0.03190844588809543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,4096,0.04096622268358866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,5120,0.02794755498568217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,5120,0.06725955671734281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,3584,0.03955022162861294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,4096,0.0540906654463874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,51200,0.6929688983493381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,3584,0.04856977860132853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,3072,0.0355057782597012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,7168,0.0734728905889723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,3072,0.019261333677503798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,2560,0.02866666515668233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,10240,0.12780711385938856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,2048,0.024113777610990737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,3072,0.043112887276543506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,1536,0.019609777463806998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,2560,0.017804443836212158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,2048,0.015934222274356417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,2560,0.03730222251680162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,1024,0.013782222237851886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,2048,0.031155556440353394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,1536,0.014477334088749357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,6144,0.08201866679721408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,1536,0.02520888878239526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,1024,0.012494222157531314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,512,0.008979555633332994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,1024,0.01905333333545261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,768,0.011737777955002256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,4096,0.023717333873112995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,256,0.006456888798210356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,512,0.011029333704047732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,768,0.016576889488432143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,512,0.013613333304723104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,256,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,64,0.004972444640265571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,128,0.010404444403118556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,32,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2560,3584,0.021533333592944678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,128,0.011368888947698804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,64,0.011037333144081963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,32,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,65536,0.535246213277181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,51200,0.3407822185092502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,768,0.01142133358452055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,16384,0.12648622194925943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,65536,0.7067706849839953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,51200,0.2434826691945394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2560,128,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,16384,0.06987644566429986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,51200,0.5563084284464518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,10240,0.082205335299174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,12288,0.056074665652381055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2560,256,0.011772444678677453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,12288,0.13487911224365234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,8192,0.06682311164008246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,10240,0.10885955227745904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,7168,0.05412355396482679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,8192,0.03853866789076064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,8192,0.08681599961386786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,6144,0.05618400043911404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,7168,0.03386755453215705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,65536,0.27727288670010036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,7168,0.07694933149549696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,6144,0.030956443813112047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,4096,0.03661866651640998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,6144,0.06702399916119046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,5120,0.025910221868091162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,3584,0.029828445778952703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,5120,0.056418663925594754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,4096,0.021036444438828364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,3072,0.02607822252644433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,4096,0.04639110962549845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,12288,0.09998844729529487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,3584,0.01943644384543101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,2560,0.023631999890009563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,16384,0.1689493391248915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,3584,0.04169866773817274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,2048,0.018017777138286166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,3072,0.018216888109842937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,2560,0.01681866745154063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,3072,0.03694044550259908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,10240,0.04936977889802721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,1536,0.014888889259762235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,2560,0.032138665517171226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,1024,0.010630221830474006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,2048,0.02693066663212246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,1536,0.01275199982855055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,768,0.00905688852071762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,1024,0.011364444262451597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,1536,0.022038223014937505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,512,0.0069857777820693115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,1024,0.016903110676341586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,768,0.01071111112833023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,5120,0.04171289006868998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,256,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,128,0.004683555414279302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,512,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,768,0.014431110686726041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,64,0.0041848888827694785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,256,0.009698666632175446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,512,0.012073777616024017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,128,0.009654222263230218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,256,0.010382222632567087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,128,0.010038221875826517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,64,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,2048,32,0.010015111002657149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,65536,0.35448087586296934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,65536,0.22787732548183867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,51200,0.3332017792595757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,2048,2048,0.015524443652894763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,51200,0.1812035507626004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,65536,0.5558506647745768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,16384,0.05988622374004788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,12288,0.08510666423373753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,51200,0.435681766933865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,10240,0.06910133361816406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,16384,0.146415114402771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,8192,0.05493866735034519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,10240,0.0410773323641883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,12288,0.10576444201999241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,10240,0.08911910984251235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,8192,0.03131644593344794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,2048,32,0.004519999855094486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,6144,0.04140799906518724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,8192,0.07335111167695788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,7168,0.02882044514020284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,7168,0.06326755550172594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,6144,0.025245333711306255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,4096,0.027772444817754958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,6144,0.05542844533920288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,5120,0.02199466692076789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,5120,0.046719110674328275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,3584,0.024009777439965144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,4096,0.018930666976504855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,16384,0.10570310884051853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,3072,0.020213334096802604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,4096,0.03874666823281182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,3584,0.01756355497572157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,2560,0.01757333344883389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,3072,0.016037333342764113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,3584,0.03478666808870103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,2048,0.015032000011867948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,2560,0.014864888456132678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,3072,0.03074666526582506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,1536,0.011804444094498953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,12288,0.04734488990571764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,1024,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,2048,0.013810666898886362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,1536,0.011790222591824003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,2048,0.02237333357334137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,768,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,7168,0.046550220913357206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,512,0.006513777706358168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,1024,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,1024,0.014508444401952954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,768,0.010326222413116032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,256,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,5120,0.03240266773435805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,512,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,768,0.012270222107569376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,128,0.004628444297446145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,64,0.0041999999019834726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,512,0.010560000108347999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,128,0.00903555585278405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,256,0.009122666385438707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,128,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,64,0.008623999853928884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,32,0.00868711123863856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,65536,0.2755875587463379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,65536,0.18098044395446777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,51200,0.20677155918545195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,2560,0.02676888969209459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,16384,0.07587022251553006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,65536,0.44599824481540257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,51200,0.1439911127090454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1536,1536,0.018761777215533786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,16384,0.049345778094397656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,51200,0.3573031160566542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,10240,0.04608711269166735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,16384,0.11574133237202962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,12288,0.03866666555404663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,8192,0.036377777655919395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,10240,0.033864001433054604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1536,32,0.004286222159862518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,7168,0.03123733401298523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,10240,0.0731511116027832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,8192,0.025608888930744592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,6144,0.027313777142100867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,8192,0.05895555681652493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,7168,0.02331022255950504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,5120,0.02293955617480808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,7168,0.05234933230612013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,6144,0.02077866593996684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,4096,0.018752000398106046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,6144,0.045759111642837524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,5120,0.018518222702874076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,3584,0.01706755492422316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,5120,0.03894666830698649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,4096,0.016178665889634024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,3072,0.014921777778201632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,4096,0.03236444460021125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,12288,0.055732442273033984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,2560,0.01278133359220293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,3584,0.02934933371014065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,3072,0.013939556148317126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,2048,0.010844444235165914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,3072,0.025976000560654536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1536,256,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,2560,0.012810666528013019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,12288,0.0867839985423618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,1024,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,2048,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,2048,0.019337778290112812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,768,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,1536,0.016182222300105624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,1024,0.009927111367384592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,512,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,1024,0.012431110772821637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,768,0.009401777552233802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,256,0.0047377778424157035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,768,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,512,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,128,0.0038453332251972626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,512,0.009395555489593083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,256,0.008636444807052612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,64,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,256,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,128,0.008387555678685507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,3584,0.015095111396577625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,32,0.003530666646030214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,128,0.007530666887760162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,65536,0.20355378256903756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,32,0.0073626670572492815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,2560,0.022639999787012737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,65536,0.17792888482411703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,51200,0.1595404413011339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,1024,1536,0.01087288889620039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,16384,0.057077334986792676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,65536,0.3667999903361003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,12288,0.04439466529422336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,51200,0.1418293317159017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,16384,0.04844889044761658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,10240,0.03484355409940084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,51200,0.2924373414781359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,16384,0.09629778067270915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,12288,0.037680887513690524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,8192,0.029370665550231934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,12288,0.07248089048597547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,10240,0.03386844529045953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,8192,0.0251893334918552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,6144,0.02198666665289137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,8192,0.05036355720625984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,7168,0.04409689042303297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,5120,0.01923377811908722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,6144,0.038682666089799665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,1024,64,0.007560888926188151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,4096,0.016082665986484952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,1024,1536,0.009169778062237633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,5120,0.01830666595035129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,3584,0.014568888478808932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,4096,0.01610933409796821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,4096,0.02723199956946903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,3072,0.012833777401182385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,3584,0.015085332923465304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,3584,0.02481777800454034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,2560,0.01076533314254549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,3072,0.013796444568369122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,7168,0.02514133354028066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,2048,0.009299555586444007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,10240,0.060953776041666664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,2560,0.012778666284349231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,7168,0.023018666439586218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,1536,0.007924444145626491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,2048,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,6144,0.02102577851878272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,1024,0.0063440000845326324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,1536,0.01074399964676963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,768,0.005470222069157495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,1536,0.013808888693650564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,1024,0.009786666267448002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,1024,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,5120,0.03313333458370633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,512,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,768,0.009393778112199571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,768,0.009673777553770278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,256,0.004181333300140169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,512,0.008974221845467886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,128,0.0034862222770849862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,256,0.008519111408127679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,256,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,64,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,768,128,0.008615111311276754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,768,32,0.0031537777847713898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,128,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,3072,0.02199111051029629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,2560,0.019256888164414298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,64,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,65536,0.18659999635484484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,32,0.006918222126033571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,2048,0.016602666841612924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,51200,0.13974755340152317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,16384,0.05016977919472588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,65536,0.31889867782592773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,51200,0.10682221915986802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,12288,0.035291555855009295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,16384,0.037393778562545776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,51200,0.2526693344116211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,16384,0.0844808883137173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,12288,0.029475556479560003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,8192,0.02277066641383701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,12288,0.06406399938795301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,10240,0.025567111041810777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,7168,0.020296888218985665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,10240,0.05338133374849955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,8192,0.018941332896550495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,6144,0.01738133364253574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,8192,0.04372622238265144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,7168,0.017249777913093567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,5120,0.015066666735543145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,7168,0.03917333483695984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,6144,0.015809777710172866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,4096,0.01293777757220798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,6144,0.03410666518741184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,5120,0.014087110757827759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,3584,0.011679111255539788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,5120,0.02931200133429633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,65536,0.13300266530778673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,3072,0.010059555371602377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,4096,0.012576888832781049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,2560,0.008851555486520132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,3584,0.011767999993430244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,3072,0.011114666859308878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,3584,0.02218933403491974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,2048,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,10240,0.02870755394299825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,2560,0.01056977775361803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,1536,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,2048,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,2560,0.017298666967286002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,1024,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,1536,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,2048,0.015088889333936902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,768,0.004546666724814309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,1536,0.012720000412729053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,1024,0.008895999855465358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,512,0.003968888686762916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,1024,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,768,0.00851733320289188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,256,0.003169777699642711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,768,0.009372444616423713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,512,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,128,0.002834666727317704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,512,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,256,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,64,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,512,32,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,256,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,512,128,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,128,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,65536,0.11564800474378799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,64,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,4096,0.02440533373090956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,51200,0.0918382207552592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,65536,0.13221422831217447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,16384,0.03237955437766181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,3072,0.019675556156370375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,65536,0.27603644794887966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,12288,0.022639110684394836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,16384,0.0369031098153856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,51200,0.21665688355763754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,10240,0.020043555233213637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,768,512,0.008369777765538957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,16384,0.07236088646782769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,8192,0.016102222932709586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,12288,0.05503466725349426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,7168,0.014624888698259989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,8192,0.018785778019163344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,10240,0.04639200038380093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,8192,0.03794222076733907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,6144,0.01314399970902337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,7168,0.017272889614105225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,7168,0.034380445877710976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,5120,0.01144088887506061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,6144,0.015678221980730694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,6144,0.03035555614365472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,4096,0.009543110926946005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,5120,0.014125333891974555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,5120,0.025755556093321905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,3584,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,4096,0.01262399968173769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,51200,0.10621155632866754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,4096,0.021433777279324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,3072,0.007867555651399825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,3584,0.01180266671710544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,12288,0.028706666496064927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,3584,0.019666666785875957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,10240,0.025460445218616064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,3072,0.01755644381046295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,2048,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,2560,0.010685332947307162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,1536,0.005519999812046687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,2560,0.015525332755512662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,2048,0.010082667072614035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,1024,0.004544888933499654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,2048,0.013464888764752282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,1536,0.009364444348547194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,768,0.0038559999730851916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,1536,0.01147377739350001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,1024,0.00869511150651508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,512,0.0032720000793536506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,1024,0.009382222261693742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,768,0.008567111359702216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,256,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,768,0.008189332981904348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,512,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,128,0.0024871110088295406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,512,0.0069857777820693115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,256,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,64,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,256,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,128,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,2560,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,128,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,256,3072,0.0110880004035102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,65536,0.07224355803595649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,64,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,51200,0.059021333853403725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,256,32,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,65536,0.13238933351304796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,51200,0.10593066612879436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,65536,0.23692001236809623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,16384,0.03695199886957804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,51200,0.18563022878434923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,16384,0.0624248915248447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,10240,0.01389422184891171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,12288,0.02803022331661648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,512,32,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,8192,0.012758221891191272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,12288,0.047636446025636464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,10240,0.02682133350107405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,7168,0.011446221835083432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,10240,0.040384001202053495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,8192,0.018922666708628338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,8192,0.03289866778585646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,7168,0.01749955614407857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,5120,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,256,32,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,7168,0.02944799926545885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,6144,0.0158942225906584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,4096,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,6144,0.02609955602222019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,5120,0.01442577772670322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,3584,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,5120,0.022333333889643352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,3072,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,16384,0.024151111642519634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,3584,0.011796444654464722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,2560,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,3584,0.017502221796247695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,3072,0.011316444310877057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,12288,0.016203555795881484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,2048,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,3072,0.015830222103330824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,2560,0.01088711122671763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,2560,0.013878222141000958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,2048,0.010024888647927178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,1024,0.0035413333939181436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,2048,0.012317332956526013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,1536,0.009480000370078618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,768,0.0032017777363459268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,1536,0.010481778118345471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,1024,0.008721777962313758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,6144,0.00999466660949919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,512,0.0028444443725877335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,1024,0.008568888737095727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,768,0.008557333714432186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,256,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,768,0.007717333734035492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,512,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,128,0.0021768889079491296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,256,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,512,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,64,0.0021768889079491296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,256,0.005628444254398346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,128,0.007589333587222629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,32,0.0021626665774318907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,4096,0.01890222231547038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,128,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,65536,0.07213955455356173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,64,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,51200,0.053322666221194796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,16384,0.019205333458052743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,128,32,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,12288,0.0118222220076455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,128,1536,0.0042160000238153665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,10240,0.010690666735172272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,8192,0.00963199966483646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,65536,0.2368160088857015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,16384,0.062349332703484424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,7168,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,12288,0.04754933383729723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,6144,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,10240,0.03993955585691664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,5120,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,8192,0.03282577792803446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,4096,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,3584,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,7168,0.02902311086654663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,6144,0.025762667258580525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,3072,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,5120,0.02232622272438473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,4096,0.018939556346999276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,3584,0.017210667332013447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,2048,0.004512888689835866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,1536,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,3072,0.01567911108334859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,2560,0.013895110951529609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,2048,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,1536,0.010437332921557956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,1024,0.008720888859695857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,512,0.002833777831660377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,51200,0.18585067325168184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,768,0.00777599960565567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,256,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,128,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,512,0.006656888872385025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,64,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,256,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,32,0.0021359999146726397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,128,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,2560,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,65536,0.07073599762386747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,51200,0.055119110478295215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,64,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3073,128,4096,0.01315199997689989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,16384,0.01718844473361969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,64,32,0.005052444421582752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,12288,0.010640888578361936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,10240,0.009709333380063375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,1024,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,8192,0.008533333738644918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,7168,0.008052444292439355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,16384,0.06189688709047106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,64,768,0.0029680000411139596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,6144,0.007669332954618666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,10240,0.040088001224729754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,5120,0.011384888655609555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,8192,0.03264000018437704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,4096,0.009685333404276106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,3584,0.008711111214425828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,7168,0.02918044394916958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,3072,0.007625777688291337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,6144,0.02551733288500044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,2560,0.007035555938879649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,5120,0.022304000125990972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,4096,0.018912000788582694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,2048,0.004459555364317364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,3584,0.017204445269372728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,1536,0.0038657776183552216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,1024,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,3072,0.015537778536478678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,768,0.0028328889360030494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,2560,0.013785777820481194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,512,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,2048,0.012088889049159156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,1536,0.01038577738735411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,128,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,768,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,64,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,512,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3073,32,32,0.0019484443797005548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,256,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,128,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,51200,0.18548799885643852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,12288,0.04763911167780558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,64,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,32,0.0051893335249688895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,1024,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3073,32,65536,0.23515023125542533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,16384,3.8867242601182728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,16384,1.2438195546468098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,16384,4.2947714063856335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,51200,4.7000304328070746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,16384,3.746406979031033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,12288,2.886367162068685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,12288,0.8551715744866265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,12288,3.1795145670572915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,10240,0.7676631079779731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,10240,2.555260340372721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,12288,2.761705822414822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,10240,2.4839901394314237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,8192,0.6084639761182997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,10240,2.2973999447292752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,8192,2.033749262491862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,8192,1.9732986026340062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,51200,11.894833034939237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,8192,1.8830267588297527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,7168,0.5364595519171821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,7168,1.830508550008138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,7168,1.658110194736057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,6144,1.4369839562310114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,6144,0.5421679814656576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,7168,1.6210337744818792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,6144,1.4250595304701064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,5120,1.2900044123331706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,5120,0.4044640064239502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,6144,1.3164230982462566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,4096,0.8829190995958117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,5120,1.1919404135810006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,4096,0.328584008746677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,4096,0.9625369177924262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,5120,1.1332079569498699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,3584,0.2954213354322645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,3584,0.8079191313849555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,4096,0.8694693247477213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,3584,0.8266515731811523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,3584,0.7575022379557291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,3072,0.2656959957546658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,2560,0.5138488875495063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,3072,0.7339430914984809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,3072,0.6647635565863715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,2560,0.22826933860778809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,2048,0.45279375712076825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,2560,0.6386088795132107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,3072,0.7288159794277616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,2560,0.5905973116556803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,2048,0.2061093383365207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,1536,0.344418658150567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,2048,0.46216975318060977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,1536,0.15584088696373835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,2048,0.49918402565850156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,1536,0.3895368840959337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,1024,0.245048893822564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,768,0.19939022594028047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,1536,0.3718577755822076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,1024,0.12633955478668213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,1024,0.3004213439093696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,512,0.15619556109110513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,768,0.11432800028059219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,1024,0.27502134111192494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,768,0.28337423006693524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,256,0.10101066695319282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,512,0.1006951133410136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,768,0.25613776842753094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,512,0.2010648912853665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,128,0.08458844158384536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,512,0.1834871106677585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,256,0.08760799964269002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,256,0.18447466691335043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,64,0.06970755259195964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,256,0.1421333286497328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,65536,128,0.08433422115113999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,51200,19.46754625108507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,128,0.16990578174591064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,128,0.1093439989619785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,64,0.17364799976348877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,65536,51200,12.999199761284721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,65536,32,0.17451911502414277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,65536,32,0.07149510913425021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,51200,3.5908232794867625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,65536,4.82493421766493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,51200,9.435870700412327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,16384,3.670881059434679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,16384,3.451360914442274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,16384,0.9665128919813367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,65536,11.974950154622396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,12288,2.2552106645372176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,12288,0.7113422287835015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,12288,2.616182327270508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,16384,2.8523858388264975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,51200,9.672684563530817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,51200,14.98677232530382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,10240,0.5695724487304688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,10240,1.9507083892822266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,12288,2.126024034288194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,10240,2.158601760864258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,8192,0.4748524559868707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,8192,1.5844550662570531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,10240,1.8086596594916449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,8192,1.6455386479695637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,7168,1.4086533652411566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,8192,1.4462418026394313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,65536,13.114951239691841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,7168,1.3258515463935006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,7168,1.1381609174940321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,6144,1.1200204425387912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,65536,18.63604566786024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,7168,0.4141226609547933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,6144,0.382022221883138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,6144,1.1400506761338975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,5120,0.3200773398081462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,5120,0.8398373391893176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,4096,0.7382648785909017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,4096,0.26237421565585667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,4096,0.7827644348144531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,5120,0.9947581821017795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,6144,1.0587422053019206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,3584,0.6278399891323513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,4096,0.6927742428249783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,5120,0.8576533529493543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,3584,0.24104266696506074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,3072,0.531417793697781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,3584,0.7283822165595161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,3072,0.2112791140874227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,2560,0.4946133295694987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,3072,0.6048515637715658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,2560,0.1797928942574395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,2048,0.3582497702704535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,2560,0.4938995573255751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,3072,0.524436420864529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,3584,0.6038968827989366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,2048,0.15255022048950195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,2560,0.4716800053914388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,1536,0.27925154897901744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,1536,0.12586132685343424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,2048,0.3648684554629856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,2048,0.4680195384555393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,1536,0.3113679885864258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,1024,0.0998835563659668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,768,0.16000444359249538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,1024,0.24536000357733834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,512,0.10813866721259223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,1024,0.24254933993021646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,768,0.0883457793129815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,768,0.19084355566236708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,256,0.08069066868888007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,512,0.16128178437550864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,768,0.17840088738335502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,512,0.15908977720472547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,256,0.14208622773488364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,256,0.06984444459279378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,256,0.11350844966040717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,64,0.057631108495924205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,1024,0.18723555405934653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,128,0.13606845008002386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,32,0.05752711163626777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,128,0.08679644266764323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,51200,1536,0.30079465442233616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,64,0.13644888665941027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,51200,32,0.138227555486891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,512,0.08022222254011366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,51200,128,0.06250222524007161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,51200,128,0.06819111108779907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,65536,1.490677303738064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,51200,1.0625066757202148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,16384,1.2238470713297527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,51200,3.0910364786783853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,65536,3.8832355075412326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,16384,1.174210654364692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,16384,0.3068924480014377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,12288,0.8202853202819824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,12288,0.8088320096333822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,16384,0.9400115542941623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,12288,0.23521155781216094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,12288,0.6658986409505209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,51200,4.793599870469835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,51200,3.346104939778646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,10240,0.19860266314612496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,10240,0.6184897952609593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,8192,0.5053813192579482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,65536,6.238131629096137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,65536,4.181544833713107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,10240,0.6954195764329699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,8192,0.17965688970353869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,10240,0.5433617697821723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,8192,0.5785786840650771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,7168,0.14657155672709146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,8192,0.44727468490600586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,7168,0.5164151191711426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,6144,0.395447121726142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,6144,0.12779643800523546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,7168,0.46379910575018984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,5120,0.28450576464335126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,7168,0.4316408899095323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,6144,0.3857742150624593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,5120,0.11073866817686294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,5120,0.4378800127241347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,4096,0.08909777800242107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,5120,0.31598933537801105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,3584,0.21076888508266875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,4096,0.276077323489719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,4096,0.23122665617201063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,3584,0.08157155248853895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,3584,0.21559555000729033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,3584,0.2164986663394504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,3072,0.07145866420533922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,2560,0.15899733702341715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,3072,0.18991466363271078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,3072,0.18341333336300322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,2048,0.12732444869147405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,2560,0.06403466727998522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,4096,0.23256444931030273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,2560,0.1614168882369995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,1536,0.08894844187630548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,2560,0.1531253390842014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,2048,0.05526933405134413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,2048,0.1333262258105808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,1024,0.06358933448791504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,2048,0.12770756085713705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,3072,0.17138755321502686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,1536,0.10593510998619927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,1536,0.1042133304807875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,768,0.04853599932458666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,1024,0.03837066557672288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,1024,0.08116533358891805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,512,0.03748355640305413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,768,0.035070220629374184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,768,0.06483733654022217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,768,0.06901066833072238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,256,0.02719377809100681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,512,0.03223289052645365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,512,0.05529600050714281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,6144,0.3403280046251085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,128,0.02083911167250739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,512,0.05771199862162272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,256,0.02907199992073907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,64,0.018604444132910836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,256,0.04993155598640442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,128,0.028493334849675495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,16384,32,0.018960888187090557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,128,0.04714488983154297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,128,0.034120000070995755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,16384,1536,0.04535733328925239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,64,0.047790222697787814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,16384,32,0.04844889044761658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,1024,0.0804542236857944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,16384,256,0.04453688859939575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,65536,1.0296026865641277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,51200,0.8322355482313367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,51200,2.4372836218939886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,16384,0.9733653598361545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,65536,2.8791315290662975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,16384,0.2373510996500651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,12288,0.5555120044284397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,16384,1.0204008950127494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,51200,3.405987633599175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,16384,0.7216373549567329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,12288,0.6214959886338975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,12288,0.17959377500745985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,51200,2.4724693298339844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,10240,0.44100266032748753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,65536,4.236595577663846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,10240,0.15264088577694362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,12288,0.5093875461154515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,8192,0.38618045383029515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,65536,3.229851616753472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,8192,0.12398844295077854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,10240,0.5424026913113064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,10240,0.43163734012179905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,7168,0.32210667928059894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,8192,0.4344880051083035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,7168,0.11102844609154595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,6144,0.27486666043599445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,8192,0.3478408919440375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,7168,0.40911110242207843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,5120,0.240814208984375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,6144,0.2963137891557482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,5120,0.0852115551630656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,6144,0.2880862289004856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,4096,0.1961075597339206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,5120,0.2469173272450765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,3584,0.1723466714223226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,4096,0.07129066520267062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,4096,0.19342399968041313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,3072,0.13041688336266413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,3584,0.17767377694447836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,4096,0.18369689252641466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,3072,0.05680711401833428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,3072,0.1499306625790066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,2560,0.11316622628106011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,3072,0.14478488763173422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,6144,0.09782666630215114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,7168,0.31591643227471244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,2048,0.09159377548429702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,2560,0.05041777756479052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,2560,0.12657333744896784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,1536,0.07867999871571858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,2048,0.04394400119781494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,2560,0.12116266621483697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,5120,0.2235964404212104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,2048,0.10313244660695393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,2048,0.10470222102271186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,1024,0.04710222283999125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,1536,0.03632355398601956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,3584,0.06357600291570027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,1536,0.08213777674569024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,768,0.04231999980078804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,1024,0.030990223089853924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,1024,0.0651351081000434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,3584,0.19714400503370497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,512,0.028569777806599934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,768,0.028284443749321833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,768,0.050811555650499135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,256,0.019863999552196927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,768,0.055552889903386436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,512,0.02590933276547326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,512,0.04315200116899279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,128,0.01921955578856998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,512,0.045796444018681846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,256,0.02347911066479153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,256,0.03856799999872843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,64,0.01609688831700219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,256,0.036291556225882635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,12288,128,0.023039999935362074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,12288,32,0.01569600072171953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,128,0.036461333433787026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,128,0.027887112564510767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,64,0.03683733277850681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,32,0.03712000118361579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,12288,1536,0.08330666356616549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,12288,1024,0.06354844570159912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,65536,0.8860755496554904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,51200,0.6977261967129178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,16384,0.8193759918212891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,51200,2.0238524542914496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,16384,0.7775608698527018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,65536,2.662173377143012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,16384,0.2065884404712253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,12288,0.4606293572319879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,51200,2.6960542466905384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,16384,0.5787315368652344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,12288,0.5017386542426215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,12288,0.15314755174848768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,10240,0.39473599857754177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,65536,3.710369957817925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,10240,0.13078667057885066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,65536,2.7402426401774087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,8192,0.3454773426055908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,10240,0.42859734429253477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,10240,0.3562897841135661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,8192,0.11998489167955186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,8192,0.34853243827819824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,7168,0.2916017903221978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,8192,0.28681688838534886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,7168,0.28707554605272084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,7168,0.09619910849465264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,6144,0.23513687981499565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,51200,2.0743590460883246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,7168,0.26208533181084526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,5120,0.19749777846866182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,6144,0.08374400271309747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,6144,0.3024800088670519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,12288,0.43907822502983945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,6144,0.21903822157118055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,5120,0.07214399841096666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,4096,0.15088178051842585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,5120,0.21675733725229898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,3584,0.1256542205810547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,5120,0.1863244507047865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,4096,0.16145155164930555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,3584,0.05412266651789347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,3072,0.09979289107852513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,4096,0.15651733345455593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,3584,0.14345955848693848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,2560,0.08412800232569377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,3584,0.15797689225938585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,3072,0.04810755451520284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,3072,0.1239360041088528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,2048,0.07729155487484403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,2560,0.04373066624005636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,3072,0.12138222323523627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,2560,0.10713599787818061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,1536,0.06065333551830716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,2048,0.0383200016286638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,2560,0.106949335998959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,2048,0.08963466352886623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,1024,0.04046133491728041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,1536,0.0310808883772956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,2048,0.08833599752850002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,768,0.033152000771628484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,1536,0.06991200314627753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,1024,0.025574222207069397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,512,0.02457155618402693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,1536,0.0720320012834337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,1024,0.05400800042682224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,768,0.023542222049501207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,256,0.017982222967677645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,768,0.04297511114014519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,512,0.020971554848882887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,768,0.04846044381459554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,512,0.03536799881193373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,128,0.014143110977278816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,512,0.04018844498528375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,256,0.01919644408755832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,256,0.03131466772821214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,256,0.03170488940344916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,64,0.01354311075475481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,128,0.019384000036451552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,10240,32,0.016107555892732408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,128,0.024149333437283833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,64,0.029954665237002905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,32,0.02981333269013299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,10240,1024,0.05741155809826321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,10240,4096,0.061796446641286217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,10240,128,0.029861334297392104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,65536,0.735970656077067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,65536,1.990841759575738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,51200,1.5500835842556422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,51200,0.5899093416002061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,16384,0.4960533248053656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,16384,0.18756355179680717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,16384,0.5440355406867133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,65536,2.753350151909722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,12288,0.38588799370659727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,51200,2.223629421657986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,16384,0.4934666421678331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,12288,0.12901422712537977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,10240,0.31302934222751194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,65536,2.164732403225369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,12288,0.4337315559387207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,12288,0.360642671585083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,8192,0.25312532318962944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,10240,0.3463733461168077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,51200,1.75128173828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,10240,0.30644623438517254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,8192,0.08880444367726643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,7168,0.2368791103363037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,8192,0.32681600252787274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,7168,0.07791911231146918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,8192,0.2454648812611898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,6144,0.18139910697937012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,7168,0.23051910930209688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,6144,0.06871555248896281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,5120,0.15685510635375977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,7168,0.2292515436808268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,6144,0.2397244506412082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,4096,0.11155554983350967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,6144,0.1904817819595337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,5120,0.05932977464463976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,5120,0.16562755902608237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,3584,0.11110488573710124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,5120,0.16074844201405844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,4096,0.051627556482950844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,10240,0.1109315554300944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,4096,0.13858844174279109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,3072,0.09433599975374009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,3584,0.04538222153981527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,4096,0.1425857808854845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,3584,0.11681866645812988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,2560,0.07206310828526814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,3584,0.11811110708448623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,2048,0.05877333217196994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,3072,0.10304088724984063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,2560,0.03654222355948554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,1536,0.04658133453792996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,3072,0.10390044583214654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,2560,0.08833422263463338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,2560,0.09127289056777954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,2048,0.07264977693557739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,1024,0.03619555632273356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,2048,0.07629688580830891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,1536,0.026192888617515564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,768,0.024572443630960252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,1536,0.06250400013393827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,1024,0.021538666552967493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,1024,0.044887112246619336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,1024,0.049116445912255176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,512,0.018533334136009216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,768,0.019827556278970506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,768,0.03648533423741659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,256,0.014471999473041959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,768,0.041610668102900185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,512,0.02915466825167338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,512,0.035283555587132774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,128,0.012056888805495368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,3072,0.04193866584036085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,256,0.025420443879233465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,256,0.016982222596804302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,64,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,128,0.016235555211702984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,8192,32,0.01127199994193183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,128,0.024694222542974684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,128,0.020654223031467862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,64,0.024696000748210486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,1536,0.05829333596759372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,8192,32,0.02438755498992072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,512,0.017792888813548617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,65536,0.6558186742994521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,8192,256,0.027653333213594224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,8192,2048,0.0313297775056627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,65536,1.7883697085910375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,51200,1.6317670610215929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,51200,0.5141875478956435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,16384,0.4750995635986328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,16384,0.5362879965040419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,16384,0.15036711427900526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,65536,2.7437057495117188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,16384,0.474806202782525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,65536,2.081743028428819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,12288,0.3676977687411838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,12288,0.11411378118726943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,12288,0.3550693458980984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,10240,0.29680177900526256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,51200,1.5916195975409613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,51200,2.099485397338867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,12288,0.34120978249443906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,10240,0.0986222227414449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,10240,0.3156435489654541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,8192,0.23689688576592338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,7168,0.17643822564019096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,10240,0.2824142244127062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,8192,0.2711075676812066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,7168,0.07294933663474189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,6144,0.14811733033921984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,8192,0.24584621853298613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,7168,0.20426489247216117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,5120,0.12150222725338405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,7168,0.20077421930101183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,6144,0.17562133736080596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,4096,0.11104533407423232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,5120,0.054930667082468666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,6144,0.17552444669935438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,5120,0.1511653396818373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,3584,0.09715288877487183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,5120,0.14794488747914633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,4096,0.045457777049806386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,4096,0.11786399947272407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,3072,0.0808782246377733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,4096,0.12184533807966445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,3584,0.04085066583421495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,3584,0.10381155543857151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,2560,0.06667021910349528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,3072,0.03645244571897719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,3072,0.09177155627144708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,8192,0.08070844411849976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,2048,0.053222222460640796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,3072,0.09590044286515977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,2560,0.07936266395780775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,2560,0.08323022392061022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,1536,0.039713776773876615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,2048,0.06512444549136691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,2048,0.028379556205537584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,6144,0.06460799773534139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,1024,0.029971556531058416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,2048,0.07185866435368855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,1536,0.023610666394233704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,768,0.02296977738539378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,1536,0.05227999885876974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,1024,0.020005333754751418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,1536,0.0584391090604994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,1024,0.03979466689957513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,512,0.017200888858901132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,1024,0.0454311105940077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,768,0.0185217774576611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,256,0.017786666750907898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,768,0.03276711040072971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,768,0.03901955485343933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,512,0.026314665873845417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,128,0.015243555108706156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,512,0.03276711040072971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,3584,0.10930399762259589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,256,0.015543111496501498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,256,0.02300888962215847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,64,0.013834666874673633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,7168,32,0.014486221803559197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,256,0.02529866662290361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,2560,0.03241511185963949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,128,0.021987555755509272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,7168,128,0.019263111882739596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,64,0.02223199937078688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,7168,32,0.02204622162712945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,512,0.016736888223224215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,65536,0.5946479903327094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,51200,1.0465315712822807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,7168,128,0.01571999986966451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,65536,1.4351030985514324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,16384,0.3620888921949599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,51200,0.5385680198669434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,16384,0.4089813232421875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,16384,0.14397954940795898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,51200,1.6326640446980794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,16384,0.40024532212151426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,12288,0.31095732582939994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,12288,0.10277244117524888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,65536,2.4737991756863065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,65536,1.7000675201416016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,10240,0.24311110708448622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,12288,0.29924533102247447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,12288,0.2836204369862874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,8192,0.1988666719860501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,10240,0.34075556861029727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,51200,1.2886906729804146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,10240,0.2383191055721707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,8192,0.20003110832638213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,8192,0.19275555345747206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,7168,0.0638071099917094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,7168,0.19814488622877333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,6144,0.13965688811408147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,7168,0.17000355985429552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,5120,0.10891377925872803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,6144,0.15210755666097006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,6144,0.15234222677018908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,10240,0.09824622339672512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,5120,0.04903822143872579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,4096,0.09381777710384792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,5120,0.12658488750457764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,8192,0.0807244446542528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,7168,0.17899910608927408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,3584,0.07968711190753512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,4096,0.03953866826163398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,4096,0.10298666689130996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,3072,0.07248977820078532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,4096,0.10700444380442302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,3584,0.03568889035118951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,3584,0.09149333503511216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,2560,0.05416444606251187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,3584,0.09380710787243313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,6144,0.05572977993223402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,3072,0.08028711213005914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,3072,0.08524355623457168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,2048,0.041442665788862444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,2560,0.029169778029123943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,2560,0.07207377751668294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,1536,0.03807555635770162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,5120,0.1288355588912964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,2048,0.05805066559049818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,1024,0.026327111654811438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,2048,0.06140622165468004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,1536,0.02123733361562093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,1536,0.04597066508399116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,768,0.019766221443812054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,1536,0.05077777637375725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,1024,0.01794488893614875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,1024,0.03535022338231405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,512,0.014710222681363424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,1024,0.03942044576009115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,768,0.016530666086408827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,768,0.02926844358444214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,256,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,3072,0.03233244352870517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,768,0.03425688876046075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,512,0.014803555276658801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,512,0.023297778434223596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,128,0.010021333065297868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,512,0.028441776831944782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,256,0.013971555564138623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,64,0.008700444466537898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,2560,0.07002488772074382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,256,0.02233244478702545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,128,0.01973600023322635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,6144,128,0.017880888448821176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,2048,0.025386666258176167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,64,0.01980266637272305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,32,0.019567999574873183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,65536,0.5431289143032497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,6144,256,0.020289777053727042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,51200,0.9164302614000109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,6144,128,0.013956444131003486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,65536,1.4202409320407445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,16384,0.2773137887318929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,51200,0.42349688212076825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,16384,0.39362043804592556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,65536,1.9484231736924913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,65536,1.4554435941908095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,16384,0.1237128840552436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,16384,0.3380906581878662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,12288,0.25766221682230633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,12288,0.09296177493201362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,10240,0.18833511405520967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,6144,32,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,12288,0.27263911565144855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,51200,1.4982479943169489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,8192,0.17362399895985922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,12288,0.2754604551527235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,51200,1.127587530348036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,10240,0.25329067971971303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,7168,0.14926221635606554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,8192,0.06189155578613281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,8192,0.18245422840118408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,10240,0.21966400411393908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,6144,0.1316186719470554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,7168,0.055904891755845815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,8192,0.18642577860090467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,7168,0.1675999959309896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,7168,0.15745866298675537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,5120,0.10732533534367879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,6144,0.04866222209400601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,6144,0.13211911254458958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,4096,0.08191555738449097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,6144,0.13449954986572266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,5120,0.04211555586920845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,5120,0.1197706725862291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,3584,0.07561777697669135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,5120,0.11909243795606826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,4096,0.035143110487196184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,4096,0.08897955550087823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,3072,0.061031997203826904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,3584,0.031938665442996554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,4096,0.0966728859477573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,3584,0.07899910873836942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,2560,0.051791111628214516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,10240,0.08025599850548638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,3072,0.02844711144765218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,3072,0.06877955463197497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,2048,0.04268444577852885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,3072,0.07487644751866658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,2560,0.025589333640204534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,2560,0.060011552439795605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,1536,0.03352266550064087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,2560,0.06656622224383883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,2048,0.022015111313925848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,2048,0.04956444435649448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,1024,0.024538666009902954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,2048,0.05611111058129204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,1536,0.01865688959757487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,1536,0.039840890301598444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,768,0.01789866718981001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,1536,0.04566488994492424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,1024,0.0158924443854226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,1024,0.0302800006336636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,512,0.01520088811715444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,1024,0.03596533338228861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,768,0.026072000463803608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,256,0.011619555453459421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,512,0.014646222194035849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,512,0.020555555820465088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,512,0.02571644385655721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,128,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,256,0.012895111408498553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,3584,0.08647644519805908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,256,0.018033777674039204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,64,0.007793777518802219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,256,0.020960888928837244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,128,0.013112000293201871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,5120,32,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,128,0.017218665944205392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,128,0.01718488832314809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,64,0.017234666479958426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,5120,32,0.01722666621208191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,5120,768,0.014892444014549255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,5120,768,0.0307608875963423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,65536,0.4261271158854167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,65536,0.9729813469780816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,51200,0.8254577848646375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,16384,0.28183555603027344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,51200,0.33525599373711484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,16384,0.11438932683732773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,12288,0.20654843913184273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,16384,0.30989689297146267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,65536,1.4132364061143663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,51200,1.0954230626424153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,16384,0.334257788128323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,12288,0.0780657794740465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,65536,1.288558218214247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,12288,0.23644532097710502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,51200,0.984458711412218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,8192,0.14155111047956678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,10240,0.06685155630111694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,12288,0.2418311172061496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,10240,0.19307022624545625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,7168,0.12784977753957114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,8192,0.06449955701828003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,10240,0.1925804482565986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,8192,0.14498310618930393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,6144,0.10942666398154365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,7168,0.04733511143260532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,7168,0.12877599398295084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,8192,0.16658222675323486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,5120,0.08499822351667617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,7168,0.13864533106486002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,6144,0.04615199897024366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,4096,0.06627821922302246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,6144,0.11121777693430583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,5120,0.03620533479584588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,6144,0.11993955241309272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,3584,0.05805333455403646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,5120,0.09314133061303033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,4096,0.030142221185896132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,5120,0.10362311204274495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,4096,0.07650399870342679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,3072,0.047544889979892306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,10240,0.1776897774802314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,3584,0.027814222706688776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,2560,0.041335999965667725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,3584,0.06633333365122478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,3072,0.02513422237502204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,3584,0.07564266522725423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,2048,0.03420711225933499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,3072,0.05872088670730591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,2560,0.0225262227985594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,3072,0.06712355878618029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,2560,0.050888889365726046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,1536,0.026499556170569524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,2560,0.05843289030922783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,2048,0.02015022271209293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,1024,0.01867555578549703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,2048,0.042414221498701304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,1536,0.01759111053413815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,2048,0.05055911011166043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,1536,0.03417511118782891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,768,0.015339555011855232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,1536,0.04096444447835287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,1024,0.01516355574131012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,1024,0.02644444505373637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,512,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,1024,0.0324524442354838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,768,0.013854222165213691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,768,0.02196888956758711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,256,0.009752000371615091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,768,0.027299554811583623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,512,0.0127368892232577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,512,0.01776711146036784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,128,0.00926755534278022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,512,0.023014222582181294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,256,0.01203733351495531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,256,0.015463999576038785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,64,0.008554666406578487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,4096,32,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,256,0.019560888409614563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,4096,128,0.012079999678664737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,128,0.014815110299322339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,128,0.0155102229780621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,64,0.014779556128713818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,4096,32,0.014807111687130399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,65536,0.3838648796081543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,51200,0.6830106841193305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,65536,1.061215082804362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,4096,4096,0.08531733353932698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,16384,0.29541065957811147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,51200,0.9473191367255317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,65536,1.2577760484483507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,51200,0.3065919876098633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,65536,1.099572393629286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,12288,0.18851999441782633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,16384,0.2647911177741157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,12288,0.07755021916495429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,10240,0.15873777866363525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,51200,0.8348604308234321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,12288,0.2036515606774224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,8192,0.12929599814944798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,10240,0.06397510899437799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,10240,0.1596577829784817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,8192,0.05046844482421875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,7168,0.10523466931449042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,8192,0.1463573376337687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,7168,0.11595911449856228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,7168,0.04281066523657905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,7168,0.12194044060177273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,16384,0.2564764552646213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,6144,0.08634489112430149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,16384,0.10530222124523586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,5120,0.07474311192830403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,6144,0.0372631119357215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,6144,0.09996355242199367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,12288,0.1973022222518921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,4096,0.059784889221191406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,5120,0.03233688738611009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,6144,0.10654666688707139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,3584,0.053336888551712036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,5120,0.0916515522532993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,4096,0.027448889282014634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,4096,0.06775733497407702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,3072,0.045331554280387036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,10240,0.1815635495715671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,4096,0.07720533344480726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,3584,0.025048888391918604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,3584,0.06030133035447863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,2560,0.040165334939956665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,3584,0.06789155801137288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,3072,0.022673777408070032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,3072,0.05342400074005127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,2048,0.03300711181428697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,3072,0.06108533011542427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,2560,0.020586666133668687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,2560,0.04637866550021701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,1536,0.02615822189384037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,2560,0.05394666724734836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,2048,0.018043556147151522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,2048,0.039056890540652804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,1024,0.018957333432303537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,2048,0.0461751123269399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,1536,0.015501333607567681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,1536,0.03091288937462701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,768,0.01587288909488254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,5120,0.08337777853012085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,1536,0.03728444377581278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,512,0.01290844463639789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,1024,0.023570666710535686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,1024,0.029543999168607924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,768,0.020058666666348774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,768,0.024733333124054804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,8192,0.12889599800109863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,512,0.016338666280110676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,512,0.012033777932325998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,512,0.021083555287784998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,128,0.006746666712893381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,256,0.01124622258875105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,256,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,256,0.018373333745532565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,64,0.005879111174080107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,128,0.011235555840863122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,128,0.013814222481515674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,32,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3584,128,0.014447111222479077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,64,0.013738666971524557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3584,32,0.013784888717863293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,768,0.012974222501118978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,65536,0.6191724671257867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3584,256,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,65536,0.3156195481618245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,51200,0.5697804556952583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,51200,0.24625778198242188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,65536,1.046820428636339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,16384,0.204605327712165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,65536,0.950624042087131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3584,1024,0.013746666411558786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,12288,0.13929333951738146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,51200,0.7173457675509982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,16384,0.23610133594936797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,16384,0.2346177895863851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,10240,0.12212533420986599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,12288,0.06569955746332805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,12288,0.16785955429077148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,8192,0.09791466924879287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,10240,0.057670222388373486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,10240,0.1450844473308987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,7168,0.08948000272115071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,8192,0.045476443237728544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,10240,0.14762489000956217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,8192,0.11364355352189805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,8192,0.12902755207485622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,6144,0.06291377544403076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,7168,0.036264889770083956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,7168,0.10241866774029201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,6144,0.03214044372240702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,5120,0.05803110864427355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,16384,0.09212978018654717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,6144,0.08953333563274807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,6144,0.09274399942821926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,5120,0.028459555572933618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,5120,0.07384888993369208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,5120,0.07993511358896892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,3584,0.03997600078582764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,4096,0.06013955672581991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,12288,0.17464266883002388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,4096,0.06712977753745185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,3584,0.02238666680124071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,3072,0.03381333417362637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,3584,0.05411110983954536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,3584,0.05945777893066406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,3072,0.020317332612143624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,2560,0.027383110589451257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,3072,0.048103110657797925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,3072,0.05331200030114916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,7168,0.11035466856426662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,2048,0.023895111348893907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,2560,0.018814222680197824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,2560,0.04172799984614054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,4096,0.0463137792216407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,1536,0.019846222466892667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,2560,0.046500444412231445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,2048,0.016537777251667447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,2048,0.035000890493392944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,1024,0.014109333356221518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,2048,0.04023911224471198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,1536,0.01482222146458096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,768,0.010972444381978778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,1536,0.028012444575627644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,4096,0.02402577797571818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,1024,0.012803555362754397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,1536,0.03344977895418803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,512,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,1024,0.026680888401137456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,768,0.011972444752852121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,256,0.006524444454246097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,768,0.018244443668259513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,768,0.022855111294322546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,512,0.014888889259762235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,128,0.006143110907740063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,512,0.020263110597928364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,256,0.01036977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,256,0.013052444491121503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,64,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,256,0.016812443733215332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,3072,32,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,128,0.010407999985747868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,3072,128,0.014365333649847241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,51200,0.8746879895528158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,64,0.01238044434123569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,32,0.01240711079703437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,65536,0.5398853090074327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,1024,0.021320000290870667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,65536,0.30056622293260365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,3072,512,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,51200,0.4624151123894586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,65536,0.8518337673611112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,3072,128,0.012464000119103326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,51200,0.2330995665656196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,65536,0.8817466629876031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,51200,0.6570809152391222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,16384,0.07383200195100573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,12288,0.12423377566867405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,16384,0.19837866889105904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,51200,0.6906728744506836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,10240,0.10689155260721843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,16384,0.22549067603217232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,12288,0.061627553568945996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,12288,0.14863822195265028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,8192,0.08811022175682916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,12288,0.17140444119771323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,10240,0.050308446089426674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,7168,0.08103733592563205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,10240,0.12878310680389404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,8192,0.03900533252292209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,8192,0.10047377480400933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,10240,0.1457937823401557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,6144,0.06278044647640653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,7168,0.03496355480617947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,8192,0.12128622002071804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,7168,0.08895999855465359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,16384,0.1668097840415107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,5120,0.054451555013656616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,6144,0.07836711406707764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,4096,0.04252533449067009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,5120,0.026983999543719824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,6144,0.0920213328467475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,3584,0.038521776596705117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,5120,0.06599199771881104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,4096,0.023315555519527856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,5120,0.07834755712085299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,4096,0.05252266592449612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,4096,0.06578755378723145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,3584,0.021360889077186584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,3584,0.04712444543838501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,2560,0.028784887658225164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,3584,0.058944887585110135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,3072,0.019142222073343065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,3072,0.041327999697791204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,2048,0.023988443944189284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,3072,0.05296088920699226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,2560,0.01756088932355245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,2560,0.03603555427657233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,1536,0.019932443896929424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,2560,0.04592533244027031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,2048,0.015842666228612263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,2048,0.030395554171668157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,1024,0.013888888888888888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,6144,0.031217777066760596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,2048,0.03918933206134372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,1536,0.014256889621416727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,768,0.011507555842399597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,1536,0.024375110864639282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,1536,0.03270311156908671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,1024,0.012439111040698158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,512,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,768,0.011730666789743634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,1024,0.025407110651334126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,768,0.016140444411171805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,3072,0.03267733256022135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,256,0.006680888848172293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,768,0.02197955548763275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,512,0.013079110946920184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,128,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,256,0.010344888601038191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,256,0.011687999798191918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,512,0.018928888771269057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,64,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,256,0.015572445260153877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2560,32,0.0058373332851462895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,128,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,128,0.011233777635627322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,128,0.012443555725945367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,64,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,32,0.011239111423492432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2560,7168,0.10594399770100911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,65536,0.43319111400180393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2560,1024,0.01867377758026123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,65536,0.26274222797817653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,51200,0.3388933340708415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,16384,0.13557510905795628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,65536,0.7033537758721246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2560,512,0.01072355525361167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,51200,0.2135688861211141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,12288,0.10283555587132771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,51200,0.5487688912285699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,16384,0.16938577757941353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,10240,0.08556533522076076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,12288,0.05040622088644239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,16384,0.20663732952541777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,51200,0.5843404663933648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,12288,0.12860622670915392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,8192,0.07166755861706205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,12288,0.14604266484578451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,10240,0.04537511203024122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,10240,0.10676799880133735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,10240,0.12630577882130942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,8192,0.035143110487196184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,8192,0.0868720014890035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,8192,0.10195110903845893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,6144,0.0502417782942454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,7168,0.07747822337680392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,7168,0.08983911408318414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,16384,0.06858844227261014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,6144,0.06668444474538167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,6144,0.028166221247778997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,5120,0.0417848891682095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,6144,0.07814933194054498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,4096,0.034550223085615374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,5120,0.024275556206703186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,5120,0.055905779202779136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,5120,0.0675377779536777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,3584,0.028988444142871436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,4096,0.02103999919361538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,4096,0.04614488946066963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,7168,0.0587902201546563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,4096,0.05690844191445244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,3072,0.024636444118287828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,3584,0.018815999229749043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,3584,0.04142933421664768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,2560,0.02278222143650055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,7168,0.0317520002524058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,3584,0.05162400007247925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,3072,0.01772711177666982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,3072,0.036970665057500206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,2048,0.01854311095343696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,3072,0.04477689001295301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,2560,0.015968888998031616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,1536,0.014481777946154276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,65536,0.7403724458482531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,2048,0.014831110835075378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,2560,0.039537777503331505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,1024,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,2048,0.03404088815053304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,1536,0.012703999876976013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,768,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,1536,0.022111111217074927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,1024,0.01128088848458396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,1024,0.01683111157682207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,512,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,1024,0.022338666849666174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,768,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,768,0.01421955558988783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,256,0.005643555687533484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,768,0.019539554913838703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,512,0.010038221875826517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,512,0.01204355557759603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,128,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,512,0.016827555166350473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,256,0.01036800030204985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,256,0.015169777803950839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,64,0.004587555511130227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,128,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,128,0.009995555712117089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,2048,32,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,2560,0.031983110639784075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,128,0.012102222277058495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,2048,0.027098667290475633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,64,0.009900444083743626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,2048,32,0.010053333308961656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,2048,1536,0.02773688899146186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,65536,0.19026311238606772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,51200,0.32033777236938477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,16384,0.10416355397966172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,65536,0.554807980855306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,51200,0.15238400300343832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,51200,0.4376968807644314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,65536,0.594164424472385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,2048,256,0.009673777553770278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,16384,0.05049333307478163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,12288,0.0826888879140218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,16384,0.15225422382354736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,10240,0.06113688813315498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,51200,0.4620159996880426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,12288,0.04054933455255296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,16384,0.15510222646925184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,8192,0.05100088980462816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,12288,0.11517866452534993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,10240,0.035014222065607704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,12288,0.11983733707004124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,7168,0.044906665881474815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,10240,0.09204533365037705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,8192,0.026729777455329895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,8192,0.07182933224572076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,10240,0.10243555572297837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,6144,0.03774044579929776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,65536,0.3850497669643826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,7168,0.023944000403086346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,5120,0.029248889949586656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,7168,0.06353155771891277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,6144,0.0217875556813346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,7168,0.07362400160895453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,6144,0.05492000116242302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,4096,0.024321777953041926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,6144,0.06491644514931573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,5120,0.04643377661705017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,3584,0.021513778302404616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,5120,0.05620977613661024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,4096,0.01647111111217075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,4096,0.038487109873029925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,3072,0.01826400061448415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,4096,0.04713155494795906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,3584,0.01521511044767168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,3584,0.034332444270451866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,2560,0.016955556141005624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,3584,0.04251289036538866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,3072,0.014131555954615274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,3072,0.03075466553370158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,2048,0.013952889376216464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,3072,0.03701333204905192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,2560,0.013089777694808112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,2560,0.0262435558769438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,1536,0.011409777734014722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,2560,0.03255911005867852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,2048,0.011928000383906893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,8192,0.08413422107696533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,2048,0.022491556074884202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,1536,0.011069333387745751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,2048,0.02812266680929396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,1536,0.01869511107603709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,768,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,1536,0.024209777514139812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,1024,0.014482667048772177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,1024,0.010040889183680216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,512,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,1024,0.0192284451590644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,768,0.009711110757456886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,768,0.012437333663304647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,256,0.004570666700601578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,768,0.017343110508388944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,512,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,512,0.01035644445154402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,128,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,256,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,512,0.015481778317027621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,256,0.009127111070685917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,64,0.0035457776652442086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,256,0.01350044459104538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,128,0.008902221918106079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,128,0.00868888861603207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,32,0.00388355553150177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1536,128,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1536,1024,0.008742222355471717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,32,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1536,5120,0.01921244462331136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,51200,0.21932621796925864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,65536,0.1785200039545695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,16384,0.07064444488949247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,65536,0.44602134492662215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,51200,0.14171111583709717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,51200,0.3469688892364502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,65536,0.5752746793958876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,12288,0.05231377813551161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,16384,0.11523021592034234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,16384,0.15447644392649332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,51200,0.45337597529093426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,12288,0.038549333810806274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,10240,0.0451128880182902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,12288,0.08653689093059963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,8192,0.03558489018016391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,12288,0.11821777290768093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,10240,0.03363200028737386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1536,64,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,7168,0.029740442832310993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,10240,0.07260533173878987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,65536,0.2779128816392687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,8192,0.025285333395004272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,10240,0.10086133082707723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,8192,0.05862577756245931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,7168,0.02295911146534814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,8192,0.08192977640363905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,7168,0.05195644497871399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,5120,0.02199466692076789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,6144,0.02055911057525211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,6144,0.045606222417619496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,4096,0.018440888987647164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,6144,0.06396622127956815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,5120,0.018215111560291715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,5120,0.038754665189319186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,16384,0.04913511210017734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,3584,0.01700799994998508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,5120,0.055092444022496544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,4096,0.01591111057334476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,4096,0.032510221004486084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,3072,0.014621333943472968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,3584,0.014835556348164877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,4096,0.04646755589379204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,3584,0.029285334878497656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,2560,0.01295555548535453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,3584,0.04097955425580343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,3072,0.013798221945762634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,3072,0.02591733303334978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,2048,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,3072,0.03621422251065572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,2560,0.012774222426944308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,2560,0.022524444593323603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,1536,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,2560,0.03197955422931247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,2048,0.011599111060301462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,2048,0.019539554913838703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,1024,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,2048,0.027710222535663184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,7168,0.07259733147091336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,1536,0.010689777632554373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,1536,0.01610488858487871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,768,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,1536,0.02306844459639655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,1024,0.012482666307025485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,512,0.004854222138722737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,1024,0.018959111637539335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,768,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,768,0.016735111673672993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,256,0.003999111139112049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,512,0.009020444419648912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,512,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,512,0.015160888433456421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,128,0.003526222374704149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,256,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,256,0.01309866706530253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,6144,0.026893332600593567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,64,0.003549333247873518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,128,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,1024,32,0.003539555602603488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,1024,128,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,64,0.007355555891990662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,32,0.007360888852013483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,1024,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,1024,768,0.010737777584128909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,65536,0.17672177155812582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,51200,0.16677333248986137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,65536,0.3670782248179118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,256,0.008560889297061497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,16384,0.055163555675082736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,65536,0.4396373430887858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,51200,0.14061066839430067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,12288,0.04148799843258328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,51200,0.28915733761257595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,16384,0.09592355622185601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,16384,0.04814933405982124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,10240,0.034879999028311834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,51200,0.34482312202453613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,12288,0.037470221519470215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,12288,0.07238844368192884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,65536,0.22534222073025176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,8192,0.026150221625963848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,12288,0.09219021929634942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,10240,0.03314044409328037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,10240,0.06071199973424276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,7168,0.02256444427702162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,10240,0.07840355237325032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,8192,0.024849777420361836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,6144,0.020058666666348774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,8192,0.04942489001486036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,7168,0.022542221678627863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,8192,0.06424355506896973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,7168,0.044080889887279935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,5120,0.01616533266173469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,7168,0.0571297804514567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,6144,0.02015466656949785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,6144,0.03850222296184964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,4096,0.013920888304710388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,5120,0.017877333694034152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,5120,0.03277066681120131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,3584,0.0121742222044203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,5120,0.04323200053638882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,1024,128,0.008348444269763099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,4096,0.01590222285853492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,3072,0.010544000400437249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,16384,0.1203848918279012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,3584,0.014828445182906257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,3584,0.02461333407296075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,2560,0.00941422250535753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,3584,0.03258133265707228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,3072,0.013801777528391944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,3072,0.0292559994591607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,2048,0.008172444171375698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,2560,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,2560,0.019275556008021038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,2560,0.02583911187118954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,1536,0.006704888823959563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,2048,0.011752000285519494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,2048,0.01681155628628201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,2048,0.022659555077552795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,1024,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,1536,0.010704889065689512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,1536,0.013766222529941134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,1536,0.019075555933846366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,768,0.00463377767139011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,1024,0.009761778016885122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,1024,0.011069333387745751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,4096,0.036433776219685875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,1024,0.016188444362746347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,512,0.003915555361244413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,768,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,3072,0.022118222382333543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,768,0.014516444669829475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,256,0.0033644443998734155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,512,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,512,0.008970666262838576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,512,0.01312711089849472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,128,0.00286666676402092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,256,0.008496888809733922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,256,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,256,0.012078222301271228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,64,0.00286666676402092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,768,128,0.00831733312871721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,128,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,768,32,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,128,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,4096,0.027141332626342773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,65536,0.16219200028313532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,64,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,32,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,51200,0.13342932860056558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,768,768,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,65536,0.32000356250339085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,16384,0.042005333635542125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,51200,0.10570400291019016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,65536,0.43068801032172305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,16384,0.08472622103161281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,12288,0.032130665249294706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,16384,0.03738755649990506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,16384,0.11815288331773545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,768,6144,0.049867557154761426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,10240,0.025246222813924152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,12288,0.02923822071817186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,12288,0.0636426673995124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,8192,0.019349333312776353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,10240,0.025919111238585577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,10240,0.05347111158900791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,65536,0.13194399409823948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,7168,0.017633777525689866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,10240,0.07646133502324422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,8192,0.01886755559179518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,8192,0.04342133469051785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,51200,0.2506942219204373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,6144,0.01513777838812934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,7168,0.017039111918873258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,7168,0.03916888766818576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,5120,0.013257777525318993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,7168,0.05632444222768148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,6144,0.015508444772826301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,51200,0.339462227291531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,6144,0.03429866830507914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,6144,0.049581332339180835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,5120,0.013991110854678683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,5120,0.02920000089539422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,3584,0.010023111270533668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,5120,0.04246133234765795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,4096,0.01238666640387641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,4096,0.024226667152510747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,3072,0.008755555583371056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,4096,0.035050666994518705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,3584,0.011742221812407175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,3584,0.022095999783939783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,2560,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,12288,0.09081333213382298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,3584,0.032278223170174494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,3072,0.011063111325105032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,3072,0.019766221443812054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,2048,0.006835555450783835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,3072,0.02825688984658983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,2560,0.010461333725187512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,2560,0.017314665847354464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,1536,0.005970666805903117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,2560,0.025240000751283433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,8192,0.06317689021428426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,2048,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,2048,0.015066666735543145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,1024,0.004864888886610667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,2048,0.022272888157102797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,768,0.004226666771703296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,1536,0.012606222596433429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,4096,0.011327111058764987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,1536,0.018591110905011494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,1024,0.010025777750545079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,512,0.003526222374704149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,1024,0.015499555402331881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,768,0.008658666577604082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,768,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,256,0.0033644443998734155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,512,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,768,0.014132445057233175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,512,0.007860444486141205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,128,0.003160888950030009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,512,0.012631110846996307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,256,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,256,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,64,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,256,0.011069333387745751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,128,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,512,32,0.002965333354141977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,128,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,512,128,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,65536,0.09891288810306126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,64,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,512,32,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,51200,0.07953333192401461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,65536,0.13098133934868708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,1536,0.00943999985853831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,65536,0.2756551106770833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,512,1024,0.00885333369175593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,16384,0.02921688887808058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,51200,0.10494222243626912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,51200,0.2157848940955268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,12288,0.02479644450876448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,65536,0.41373777389526367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,16384,0.07196355528301664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,51200,0.32572799258761936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,16384,0.11443732844458686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,12288,0.0284186667866177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,12288,0.055191109577814736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,8192,0.017483555608325534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,12288,0.08815644184748332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,10240,0.024901333782407973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,10240,0.04658400019009908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,7168,0.013920888304710388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,10240,0.07427110936906603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,8192,0.01846311158604092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,8192,0.03796622157096863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,6144,0.013618666264745923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,8192,0.06061866548326281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,7168,0.016799110505315993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,7168,0.03414933217896356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,5120,0.010652444428867765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,6144,0.015484443969196744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,7168,0.05391200052367317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,6144,0.03001333276430766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,4096,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,6144,0.047396444612079196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,16384,0.03668177790111966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,5120,0.025566221939192876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,3584,0.007582222421964009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,5120,0.039769778649012245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,10240,0.020640000700950623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,4096,0.012411555482281579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,4096,0.021366222037209406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,3072,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,4096,0.03393244412210252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,2560,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,3072,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,3072,0.017504889104101393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,3072,0.027302222119437322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,2048,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,2560,0.0107351111041175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,2560,0.015557333827018738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,2560,0.024403555525673762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,1536,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,2048,0.010023111270533668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,2048,0.013447110851605734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,2048,0.021418665846188862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,1536,0.009382222261693742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,1024,0.004087999877002505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,1536,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,5120,0.013975999421543546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,768,0.0033297776761982176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,1536,0.017998221847746108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,1024,0.00870488915178511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,1024,0.009094222552246517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,512,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,768,0.008361777497662438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,1024,0.015150222513410779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,768,0.008055110772450766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,3584,0.011754666765530905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,256,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,3584,0.019535111056433786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,768,0.01348888874053955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,128,0.002459555657373534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,3584,0.030894223186704848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,512,0.007234666910436418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,256,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,512,0.012449777788586087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,128,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,256,0.010674667027261523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,128,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,32,0.002532444480392668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,256,128,0.00832622249921163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,65536,0.07120888763003878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,64,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,51200,0.058132443163130015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,65536,0.13030221727159289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,65536,0.23723467191060385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,51200,0.18565689192877877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,51200,0.10457066694895427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,65536,0.40964799457126194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,256,512,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,51200,0.32288000318739146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,16384,0.06226488616731432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,256,0.005980444451173146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,16384,0.03669510947333442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,16384,0.11320800251430935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,12288,0.028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,10240,0.01200711147652732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,12288,0.08686577611499363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,10240,0.04052177733845181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,10240,0.07288799683252971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,16384,0.021914665897687275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,8192,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,10240,0.025677333275477093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,8192,0.01902222302224901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,7168,0.01148088855875863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,256,64,0.002464888824356927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,8192,0.032968888680140175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,8192,0.05977333254284329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,6144,0.010281778044170803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,7168,0.017190222938855488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,7168,0.02940533227390713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,12288,0.047621332936816745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,7168,0.05377689003944397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,256,32,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,6144,0.015838222371207345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,6144,0.025964443882306416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,4096,0.007738666401969061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,6144,0.04636977778540718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,5120,0.022329777479171753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,3584,0.007124444676770105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,4096,0.012416000167528788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,5120,0.03986755675739712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,4096,0.018976888722843595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,3072,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,4096,0.033527112669414945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,3584,0.01181333346499337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,3584,0.017243555850452848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,2560,0.005633777628342311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,3584,0.03050400151146783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,3072,0.011065777805116443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,3072,0.01578044394652049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,12288,0.013768889009952545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,2048,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,3072,0.027102222045262653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,2560,0.010533333652549319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,2560,0.013826666606797112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,1536,0.004238222208287981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,2048,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,2048,0.012358222570684222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,2560,0.02404444416364034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,1024,0.003548444559176763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,5120,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,1536,0.009452444811662039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,1536,0.01038755559258991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,1536,0.017853332890404593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,1024,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,5120,0.014145778285132514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,1024,0.008403555386596257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,512,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,1024,0.015004444453451367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,768,0.008379555410808986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,768,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,256,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,512,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,512,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,768,0.013101333545313941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,128,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,512,0.011766222616036734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,256,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,64,0.0021955555097924336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,256,0.010284444524182213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3072,128,128,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,32,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,128,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,65536,0.06738666693369548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,51200,0.05400355656941732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,64,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,32,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,16384,0.01901333365175459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,12288,0.015614221493403116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,2048,0.020981333321995203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,65536,0.23782221476236978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,10240,0.013739555246300168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,128,768,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,8192,0.011669333610269757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,7168,0.010889777706729041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,16384,0.061786664856804736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,12288,0.047522667381498546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,6144,0.009822222093741098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,5120,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,10240,0.03990933299064636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,8192,0.032560888263914324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,4096,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,3584,0.0063591111037466265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,7168,0.02921244502067566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,6144,0.025803556044896443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,3072,0.005668444352017508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,5120,0.02235200007756551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,128,256,0.00564444437623024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,4096,0.018903111418088276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,2048,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,3584,0.017215111189418368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,1536,0.0038693332009845306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,3072,0.01571822166442871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,3072,128,128,0.00810222244924969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,1024,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,2560,0.013794666363133324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,768,0.0031315556002987754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,512,0.002587555597225825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,2048,0.012176888684431711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,256,0.002518222149875429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,1024,0.008485332959228093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,768,0.007725333174069722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,128,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,64,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,512,0.006830222076839871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,32,0.002208000048995018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,256,0.00565422202150027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,51200,0.18567466735839844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,65536,0.06575377782185872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,128,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,64,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,32,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,64,2560,0.005149333427349727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,65536,0.2350213262769911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,12288,0.010188444621033138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,10240,0.009128000173303816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,51200,0.18514577547709146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,16384,0.062124444378746875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,8192,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,7168,0.007667555577225155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,12288,0.0475413335694207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,6144,0.0071928890215026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,10240,0.03978577918476529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,64,1536,0.01036622209681405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,5120,0.006791111081838608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,4096,0.006218666831652324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,6144,0.02562933332390255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,3584,0.007859555383523306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,5120,0.022264000442292955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,3072,0.006797333144479328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,51200,0.04962311188379923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,2560,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,3584,0.01717333330048455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,2048,0.004520888957712385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,3072,0.01552088889810774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,1536,0.003891555385457145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,2560,0.01387466655837165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,1024,0.0031911111954185697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,2048,0.012118221984969245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,768,0.002830222249031067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,1536,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,512,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,256,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,1024,0.008358221915033128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,8192,0.03273333443535699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,128,0.002161777681774563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,7168,0.028844445943832397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,768,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,64,0.0018826667219400406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,32,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,256,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,128,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,4096,0.018927110566033255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,32,0.004975111120276981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3072,32,16384,0.017278222574128043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,512,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3072,32,64,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,16384,1.1661982006496854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,12288,0.9443786409166125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,16384,3.4478382534450955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,12288,2.622004403008355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,16384,4.043565538194445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,12288,2.8328497144911022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,10240,0.763120863172743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,51200,4.482476552327474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,8192,0.6333528624640571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,10240,2.383514616224501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,8192,1.8288230895996094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,10240,2.3343626658121743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,8192,1.8143910302056208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,7168,0.5265493392944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,6144,0.4655582110087077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,6144,1.3829902013142903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,7168,1.7437298032972548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,6144,1.3274959988064237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,7168,1.6186550988091364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,5120,0.394703123304579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,5120,1.362587504916721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,4096,0.3308702309926351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,5120,1.1378036075168185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,3584,0.2887875503963894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,4096,0.8798604541354709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,4096,1.0439377890692818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,3584,0.8518542183770074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,3072,0.6425617535909017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,3584,0.8242702484130859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,3072,0.26105332374572754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,2560,0.6635066668192545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,2560,0.224488894144694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,2048,0.4561884668138292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,3072,0.7607475386725532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,2048,0.21268178357018364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,1536,0.3426240020328098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,2048,0.46299113167656797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,1536,0.15312711397806802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,1024,0.25083910094367134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,1536,0.4307057857513428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,768,0.19684444533454049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,1024,0.1212631066640218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,1024,0.27833512094285756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,512,0.15645599365234375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,768,0.12481066915724014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,51200,11.072845458984375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,768,0.22454754511515299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,512,0.09823199775483872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,512,0.18739022148980033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,128,0.07856444517771403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,256,0.08559644222259521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,2560,0.5988871256510416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,64,0.06605066855748494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,65536,128,0.0831022196345859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,32,0.06690933307011922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,128,0.1595146656036377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,256,0.1725155512491862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,64,0.16178045007917616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,32,0.16414577431148952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,65536,256,0.1146728860007392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,16384,2.7445021735297312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,51200,3.46232541402181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,65536,4.551765441894531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,16384,0.8804684215121799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,65536,51200,18.056232028537327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,16384,3.1425938076443143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,12288,0.6769253412882487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,12288,2.5589057074652777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,12288,2.4259635077582464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,10240,1.905777825249566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,51200,8.87610117594401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,10240,0.5774328973558214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,10240,2.0329600440131292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,8192,0.463955561319987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,8192,1.7116115358140733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,7168,0.4297235541873508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,8192,1.4303199979994032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,7168,1.5143689049614801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,7168,1.2468933529324002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,6144,0.36744266086154515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,65536,11.597125583224825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,6144,1.1363858116997612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,5120,0.3083244429694282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,5120,1.0419990751478407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,6144,1.0991555319892037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,4096,0.2537359926435682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,4096,0.7290648884243436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,3584,0.22794132762485078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,4096,0.7251662148369683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,3584,0.7452701992458768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,51200,12.715116712782118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,3584,0.6151946915520562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,3072,0.20198488235473633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,2560,0.17719644970364043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,3072,0.6247662438286675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,2560,0.4865733252631293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,3072,0.5421982341342503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,2048,0.3704088793860541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,5120,0.8763253423902723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,2560,0.4616977903578017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,2048,0.15790843963623047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,1536,0.3150888813866509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,1536,0.12471555338965522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,1024,0.20485688580407035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,2048,0.36631022559271914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,1024,0.09860266579522027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,1536,0.2883591122097439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,512,0.12546932697296143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,768,0.0884737769762675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,256,0.08236000272962782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,512,0.07728977998097737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,768,0.18813155757056343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,1024,0.21867288483513725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,128,0.059149331516689725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,512,0.1525004439883762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,256,0.0710773335562812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,64,0.054655111498302884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,51200,128,0.0663964417245653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,256,0.13427110513051352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,32,0.06015377574496799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,128,0.12926310963100857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,64,0.1282880041334364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,32,0.12987200419108072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,51200,768,0.16637067000071207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,51200,65536,16.297281901041668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,16384,1.0473990970187717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,51200,1.046402719285753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,65536,1.4298853344387477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,16384,0.3032097816467285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,12288,0.23113777902391222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,12288,0.8776755332946777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,16384,1.011231952243381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,12288,0.8103120062086316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,51200,3.0944656795925565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,10240,0.6045253541734483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,10240,0.1994240019056532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,8192,0.4683866500854492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,10240,0.6226542260911729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,65536,3.8325038486056857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,8192,0.16143643856048584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,7168,0.4200577735900879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,8192,0.602859550052219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,7168,0.13900978035397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,7168,0.4091368781195746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,51200,4.125301361083984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,6144,0.12395111719767253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,5120,0.3049608866373698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,5120,0.10785422060224746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,6144,0.44211732016669375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,4096,0.2504764397939046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,5120,0.29077688852945965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,4096,0.0895964437060886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,3584,0.18990755081176758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,3584,0.07926399840248956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,4096,0.3082115650177002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,3072,0.18937155935499403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,3584,0.24347821871439615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,2560,0.14144355720943874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,3072,0.07022844420539008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,3072,0.19047999382019043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,2048,0.11469688680436875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,6144,0.34288801087273496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,65536,5.545578850640191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,2560,0.06268266836802165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,1536,0.09221777651045059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,2560,0.15455110867818198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,1024,0.06252444452709623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,2048,0.05392711030112373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,768,0.04914399981498718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,2048,0.1253839996125963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,1024,0.03671111000908746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,512,0.03709955679045783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,1536,0.05076711045371162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,1536,0.09907199939092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,1024,0.09080000056160821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,256,0.026499556170569524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,512,0.031118220753139917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,768,0.06272266970740424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,256,0.028130667077170476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,128,0.027484445108307734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,256,0.04860977662934197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,128,0.044630222850375705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,32,0.020220445262061227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,64,0.045078221294615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,32,0.04582755433188545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,16384,768,0.033512890338897705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,128,0.024977778395016987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,16384,512,0.05232888791296217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,65536,1.041060447692871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,16384,0.6614053514268663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,51200,0.8141804271274142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,16384,0.23528711001078287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,16384,64,0.02049777739577823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,51200,2.1719396379258895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,65536,2.617016050550673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,12288,0.17768000231848824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,12288,0.619126213921441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,16384,0.9390142228868273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,10240,0.434550232357449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,12288,0.5565564367506239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,10240,0.15132621924082437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,8192,0.39094310336642796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,8192,0.13548621866438124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,10240,0.4740533298916287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,7168,0.29124267896016437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,8192,0.35699910587734646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,7168,0.12147911389668782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,6144,0.2795235580868191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,7168,0.31239377127753365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,6144,0.09592622518539429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,5120,0.22301954693264434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,6144,0.2732844352722168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,5120,0.09229777918921576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,4096,0.06874755356046888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,65536,3.9314604865180125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,51200,3.086250729031033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,3584,0.14442133903503418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,5120,0.26098312271965873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,4096,0.18172800540924072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,3072,0.13069600529140896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,3584,0.06289244360393949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,2560,0.1042239997122023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,3072,0.0555004444387224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,2048,0.09081333213382298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,3584,0.16485688421461317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,2560,0.054132444991005786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,3072,0.15267554918924967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,1536,0.06580888562732272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,2048,0.04275200102064344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,2560,0.11751377582550049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,1024,0.04768266611629062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,2048,0.0977368884616428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,768,0.0411155554983351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,512,0.029902223083708022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,1024,0.029236445824305218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,1024,0.058893336190117725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,768,0.04833155539300707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,256,0.019903999235894945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,512,0.0244159996509552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,512,0.039284444517559476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,128,0.01753777762254079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,4096,0.17138844066196016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,256,0.02139555580086178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,64,0.014736000034544202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,256,0.03548444310824076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,12288,32,0.014813333749771118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,128,0.021366222037209406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,128,0.03324000040690104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,64,0.0332951115237342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,32,0.03358133302794562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,1536,0.03663999835650126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,12288,1536,0.07775288820266724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,12288,768,0.02681511143843333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,16384,0.5945173369513618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,65536,0.870783117082384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,51200,0.6841466691758898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,16384,0.25715555085076225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,51200,1.7903387281629775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,16384,0.6481351322597927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,12288,0.19971466064453125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,65536,2.388522677951389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,12288,0.4499084684583876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,10240,0.3608666790856256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,10240,0.13013511233859593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,8192,0.3025137848324246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,10240,0.3758480019039578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,7168,0.28867554664611816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,8192,0.11219733291202122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,8192,0.2978106604682075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,7168,0.09196266863081191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,6144,0.24534577793545195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,6144,0.08160710997051664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,7168,0.3096880118052165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,5120,0.19208978282080758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,6144,0.2373110983106825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,4096,0.13376800219217935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,5120,0.06995377937952678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,12288,0.5745386547512478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,3584,0.11796532736884223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,4096,0.05897511376274956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,5120,0.2474515438079834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,3072,0.10376889175838894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,3072,0.048004445102479726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,3584,0.13271466890970865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,2560,0.08599199851353963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,2048,0.06833955314424303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,3072,0.11883021725548638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,2560,0.0418471097946167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,1536,0.05607555309931437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,2048,0.03601866629388597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,2560,0.10087022516462539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,1024,0.04393511017163595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,2048,0.08324888679716322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,1536,0.029527111185921565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,51200,2.692047966851128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,768,0.03406577640109592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,1536,0.06624888711505465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,1024,0.024504888388845656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,512,0.024579554796218872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,1024,0.05032622151904636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,65536,3.4427146911621094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,768,0.022268444299697876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,256,0.01754666699303521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,512,0.02013511127895779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,128,0.014541332920392355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,768,0.040714667903052434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,64,0.01311644415060679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,512,0.03297955460018582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,256,0.018415111634466384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,10240,32,0.0170462214284473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,128,0.01850133306450314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,256,0.02905422117975023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,128,0.028154667880800035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,64,0.027859555350409612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,32,0.027775110469924078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,65536,0.7522977723015679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,51200,0.5790978007846409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,10240,4096,0.1609119971593221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,16384,0.1619244416554769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,51200,1.545696046617296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,16384,0.6007857852511936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,65536,1.8682053883870442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,16384,0.5703031222025553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,12288,0.14764977826012507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,10240,0.2974177731408013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,12288,0.3790160020192464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,10240,0.12858045101165771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,51200,2.10427369011773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,10240,0.3211333221859402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,8192,0.22908266385396323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,10240,3584,0.05293155709902445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,8192,0.0891466670566135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,7168,0.2303395536210802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,65536,2.687664879692925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,8192,0.24603732426961264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,6144,0.2020933363172743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,6144,0.07614133093092176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,7168,0.2168337768978543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,5120,0.15970667203267416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,6144,0.18824444876776802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,4096,0.11095199982325236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,5120,0.06342133548524645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,3584,0.09530755546357896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,5120,0.15588533878326416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,12288,0.3713840113745795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,4096,0.12390844027201335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,3072,0.0895306666692098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,3584,0.04371822211477491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,2560,0.07415022452672322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,3584,0.11629154947068955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,3072,0.03883822096718682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,2048,0.06678844160503812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,3072,0.09624799754884507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,2560,0.03453600075509813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,1536,0.04990488953060574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,2560,0.08235910865995619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,2048,0.0304115547074212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,1024,0.03477688961558872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,2048,0.06896266672346327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,7168,0.07652178075578478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,1536,0.025257777836587694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,768,0.024902222885025874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,1536,0.05428533421622383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,512,0.019885333047972787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,1024,0.020824889341990154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,256,0.014519110321998596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,768,0.018983110785484314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,1024,0.041690667470296226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,512,0.017369776964187622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,128,0.01237155579858356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,768,0.03438222077157762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,512,0.02777422302299076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,256,0.02453244394726223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,4096,0.048728889889187284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,128,0.023317333724763658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,64,0.023357333408461675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,8192,32,0.023249778482649062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,64,0.010399111443095738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,128,0.01590844492117564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,65536,0.6461324691772461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,8192,32,0.01069155583779017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,65536,1.7266107135348852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,51200,1.2675946553548176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,51200,0.5084720187717015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,16384,0.4469715754191081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,16384,0.15104356076982287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,12288,0.1126755608452691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,12288,0.3256782160864936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,8192,256,0.01571733256181081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,12288,0.4436631202697754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,10240,0.10413066546122234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,8192,0.1997742255528768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,65536,2.6345110999213324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,7168,0.1806186702516344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,51200,1.8065253363715277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,16384,0.47458934783935547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,8192,0.22353866365220812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,7168,0.0678222245640225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,6144,0.16642133394877115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,5120,0.13391289446089003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,6144,0.06201510959201389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,4096,0.11578578419155544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,5120,0.05156711075041029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,6144,0.165011551645067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,5120,0.1475600004196167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,10240,0.27041512065463597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,4096,0.04352444410324097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,3072,0.0839502215385437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,4096,0.11186933517456055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,3584,0.03930844532118903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,10240,0.27217067612542045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,2560,0.06315022044711642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,3584,0.10003199842241074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,3072,0.03731288843684726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,2048,0.05469155642721388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,3072,0.08580622408125137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,2560,0.03196355700492859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,1536,0.03762044509251913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,2048,0.027284445034133062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,2560,0.07408266597323947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,1024,0.027718222803539697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,7168,0.20841599835289848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,768,0.021341333786646526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,1536,0.049068444304996066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,1024,0.019137778215938143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,512,0.016400888562202454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,1024,0.03769066598680284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,768,0.017666666044129264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,256,0.012905778156386482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,768,0.030643555853101943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,512,0.016171556380059984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,3584,0.08266755607393053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,128,0.00998488896422916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,512,0.025036444266637165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,256,0.01484266585773892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,256,0.021916444102923077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,64,0.009000889129108852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,128,0.015067555838161044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,7168,32,0.009751111268997192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,128,0.020971554848882887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,8192,0.08004888561036852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,64,0.021115556359291077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,32,0.020990222692489624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,7168,2048,0.061365332868364125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,7168,1536,0.02316177719169193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,16384,0.3274799982706706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,65536,0.6027741962009007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,51200,0.48479202058580184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,51200,0.9509378009372287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,65536,1.3913467195298936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,16384,0.1538497739368015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,12288,0.2921644316779243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,16384,0.38099998897976345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,12288,0.11319377687242295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,12288,0.3032471073998345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,51200,1.4799458185831706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,10240,0.30880533324347603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,8192,0.07099999984105428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,65536,2.0200862884521484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,8192,0.19154932763841417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,6144,0.13256088892618814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,7168,0.0624248915248447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,7168,0.16691466172536215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,5120,0.11394666300879584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,6144,0.057257778114742704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,6144,0.14508533477783203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,4096,0.08226578103171454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,5120,0.04756711257828606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,3584,0.07814400063620673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,4096,0.03911644551489089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,5120,0.1212186680899726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,4096,0.09739644659890069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,10240,0.0873253345489502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,3072,0.06033777528338962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,3584,0.03507555524508158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,8192,0.18235822518666586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,3072,0.03178222311867608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,3584,0.08613422181871201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,2048,0.04556711183653938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,2560,0.028231110837724473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,3072,0.07579733265770806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,1536,0.03435644507408142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,2560,0.0654275549782647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,7168,0.16481599542829725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,1024,0.024350222614076402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,2048,0.054642667373021446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,1536,0.02088711162408193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,768,0.019160888261265226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,1536,0.04349422123697069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,512,0.014338667194048563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,1024,0.03321511215633816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,768,0.016043555405404832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,256,0.010614222122563256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,768,0.02755022214518653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,512,0.014789332946141561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,128,0.009567110902733272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,512,0.022310222188631695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,256,0.013615111509958902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,64,0.008086221913496653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,256,0.019483556350072224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,32,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,128,0.013784888717863293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,128,0.01862844493654039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,2560,0.05008000135421753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,64,0.018890667292806838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,6144,32,0.018592000007629395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,2048,0.024689777029885188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,6144,1024,0.017512889371977914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,65536,0.5305635664198134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,16384,0.27501244015163845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,51200,0.41739821434020996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,65536,1.1033111148410373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,6144,10240,0.23858756489223906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,51200,0.9616649415757922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,16384,0.12358933024936253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,12288,0.2553893460167779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,10240,0.18888444370693633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,12288,0.09047377771801418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,16384,0.343303124109904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,8192,0.15775199731191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,12288,0.25250310368008083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,10240,0.0823813345697191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,8192,0.06352888875537448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,51200,1.3124986224704318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,7168,0.13926045099894205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,10240,0.20214933819240996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,8192,0.16055999861823186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,7168,0.05714133050706652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,65536,1.8608105977376301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,5120,0.09500000211927627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,6144,0.05397866831885444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,4096,0.08212977647781372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,7168,0.14401777585347494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,5120,0.041583110888799034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,3584,0.06830844614240858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,4096,0.03715466790729099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,5120,0.10354666577445136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,3072,0.057610664102766246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,3584,0.030978666411505804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,3584,0.09098488754696316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,2560,0.04597155584229363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,3072,0.029042667812771265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,3072,0.06507466899024116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,2048,0.036920888556374445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,2560,0.025055999557177227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,1536,0.028493334849675495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,2560,0.058562669489118785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,2048,0.021549334128697712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,1024,0.020098666350046795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,2048,0.04706311225891113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,1536,0.018226666582955253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,768,0.016709332664807636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,6144,0.11823911137051052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,1024,0.015298666225539314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,1024,0.029080000188615587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,512,0.012818666795889536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,768,0.014780445231331719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,256,0.009328000247478485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,512,0.013405332962671915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,512,0.019380443625979953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,128,0.008303999900817871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,4096,0.08419644170337254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,256,0.017210667332013447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,64,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,128,0.012475555141766867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,5120,32,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,128,0.016532444291644625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,64,0.016522667474216886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,32,0.0163102216190762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,1536,0.0381004446082645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,6144,0.1325297753016154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,65536,0.42020355330573184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,5120,256,0.012548444171746572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,65536,0.9897795783148872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,51200,0.616703987121582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,16384,0.24202577273050943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,12288,0.17519377337561715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,16384,0.1264844470553928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,51200,0.3224284383985731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,16384,0.2654213375515408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,10240,0.2075288825564914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,12288,0.08419377936257257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,65536,1.288100454542372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,12288,0.22227554851108125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,8192,0.1450773345099555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,10240,0.1705706649356418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,7168,0.11697155899471706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,8192,0.053914666175842285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,51200,1.037001821729872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,8192,0.1445715559853448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,5120,768,0.024453333682484094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,7168,0.0468444459968143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,5120,0.07695021894243029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,6144,0.04306755463282267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,7168,0.12807644738091364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,4096,0.06226311127344767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,5120,0.035587555832333036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,6144,0.10587555832333034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,3584,0.055480887492497764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,5120,0.08643644385867649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,3072,0.04670222269164192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,3584,0.027054222093688116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,4096,0.0716257757610745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,2560,0.040087110466427274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,3584,0.0631564458211263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,3072,0.024804444776640996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,2048,0.0335804455810123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,3072,0.05487111210823059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,2560,0.021913778450753953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,1536,0.025398222936524287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,2560,0.048231111632453076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,2048,0.01935822268327077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,1024,0.016359110673268635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,2048,0.040243556102116905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,1536,0.017049777838918898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,10240,0.06452977657318115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,768,0.013264888690577613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,512,0.010802666346232096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,1024,0.01447111037042406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,768,0.013451555536852943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,1024,0.02461066676510705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,256,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,6144,0.08837599886788262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,128,0.0063386667105886675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,512,0.01679733395576477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,256,0.011634666886594562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,64,0.005653333332803514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,256,0.01480977733929952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,128,0.011730666789743634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,4096,32,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,4096,0.03085244364208645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,128,0.01420088940196567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,64,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,32,0.014335110783576965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,65536,0.3744408819410536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,1536,0.0322693321439955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,51200,0.5881431367662218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,65536,0.9938817554050021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,4096,512,0.01239111108912362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,16384,0.21490666601392958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,16384,0.0893991125954522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,51200,0.2990186744266086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,12288,0.1586186620924208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,65536,1.3071084552341037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,12288,0.06880444288253784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,10240,0.13286133607228598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,12288,0.1775733364952935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,10240,0.06002489063474867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,51200,0.9528000089857313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,10240,0.18286222881740996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,7168,0.09469955497317845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,8192,0.04664444592263964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,8192,0.12071200211842854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,6144,0.08779111173417832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,7168,0.04493955439991421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,4096,768,0.020828444096777175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,5120,0.06974311007393731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,7168,0.10957155625025432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,6144,0.036574221319622464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,4096,0.05935733185874092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,6144,0.09234844313727485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,3584,0.0473271111647288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,4096,0.02715911136733161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,5120,0.07823733488718669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,3072,0.04141155547565884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,3584,0.024450666374630396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,3584,0.0570346646838718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,8192,0.11528800593482123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,2560,0.03215555681122674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,3072,0.022064889470736187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,2048,0.027719111906157598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,3072,0.05043288734224108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,1536,0.022037333912319605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,2560,0.020004444652133517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,2048,0.017638223038779367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,2560,0.04375377628538343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,1024,0.0162444445821974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,2048,0.036313778824276395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,1536,0.015007111761305066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,768,0.013551111022631327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,1536,0.029300444655948218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,1024,0.013359111216333179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,512,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,768,0.012648888760142855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,1024,0.0229093333085378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,256,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,512,0.011591999895042844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,5120,0.03222044308980306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,128,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,512,0.015464888678656684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,256,0.011044444309340583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,64,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3584,32,0.005636444522274866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,256,0.013777777552604675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,4096,0.06351288821962145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,16384,0.33157867855495876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,64,0.01316177762216992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,32,0.013144888811641268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,65536,0.5503493414984809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,65536,0.30017511049906415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,51200,0.543217764960395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,16384,0.17102489206526014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,51200,0.2414097785949707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,768,0.019139554765489366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,16384,0.07890222469965617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,12288,0.12723999553256565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,65536,0.9957378175523547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,51200,0.7666186756557889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3584,128,0.0107004443804423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,16384,0.23790221744113496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,12288,0.06303022305170695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,10240,0.10649066501193577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3584,128,0.01314222233162986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,8192,0.09267644749747382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,12288,0.15735377205742732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,7168,0.07809778054555257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,8192,0.0407626661989424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,6144,0.07247555255889893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,7168,0.03559466534190708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,10240,0.13783021767934164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,8192,0.11258221997155084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,5120,0.057017776701185435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,6144,0.032736887534459434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,7168,0.09602222177717422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,5120,0.02831288841035631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,6144,0.09157244364420573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,3584,0.038753777742385864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,4096,0.024124445186720953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,3072,0.033018665181265935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,5120,0.06937422355016072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,4096,0.0563653310139974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,2560,0.027679110566775005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,3072,0.0199946661790212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,3584,0.051126221815745033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,2560,0.017902221944597032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,3072,0.044906665881474815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,2560,0.03909600112173293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,1536,0.01795288920402527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,2048,0.01608000033431583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,2048,0.03285422258906894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,1024,0.013528889252079858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,1536,0.014192889134089151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,10240,0.055621332592434354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,1536,0.02648799949222141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,768,0.01073688848151101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,1024,0.012142221960756512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,512,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,1024,0.01997777819633484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,768,0.011370666325092316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,4096,0.04688266581959195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,768,0.017263111140992906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,512,0.010680889089902243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,128,0.0058248887459437055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,512,0.014161777165200977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,256,0.009995555712117089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,3584,0.02232355541653103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,64,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,32,0.005468444277842839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,3072,128,0.010123555858929953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,2048,0.024303999212053087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,64,0.012081777883900536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,32,0.01203733351495531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,65536,0.615313794877794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,65536,0.28631199730767143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,51200,0.42251910103691953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,3072,256,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,16384,0.1623822185728285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,65536,0.8098319901360406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,51200,0.2267413404252794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,12288,0.10955822467803955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,256,0.012448000411192576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,10240,0.07945511076185438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,16384,0.18356888824039033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,3072,128,0.012034667034943899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,12288,0.05793955590989855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,8192,0.07934933238559298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,51200,0.6546488867865669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,7168,0.06384978029463026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,8192,0.039423998859193586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,6144,0.05836800071928236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,7168,0.0344782206747267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,8192,0.09456355704201592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,5120,0.04511911008093092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,7168,0.08611200253168742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,6144,0.030582222673628066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,4096,0.036990222003724836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,6144,0.07387022177378337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,5120,0.027228444814682007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,5120,0.06103999747170342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,4096,0.022597332795461018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,4096,0.05026311013433668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,16384,0.07187910874684651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,3584,0.02084355552991231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,3072,0.02896089024013943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,2560,0.025455110602908667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,3584,0.04513422316975064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,3072,0.01904711127281189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,2048,0.021439111895031396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,2560,0.017287999391555786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,3072,0.04018488857481215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,10240,0.048455112510257296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,1536,0.017156443662113614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,2560,0.034838222795062594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,10240,0.11777244673834907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,1024,0.012503999802801343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,2048,0.015506666567590503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,2048,0.029473778274324205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,1536,0.013717333475748697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,1024,0.012089778151777057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,512,0.00960355583164427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,1024,0.01794311073091295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,768,0.011145778000354767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,256,0.006803555621041193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,768,0.015840000576443143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,512,0.010371555884679159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,128,0.005478222337034013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,512,0.012720888687504662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,3584,0.033744888173209295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,64,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,256,0.009953777823183272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,32,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,256,0.011295110815101199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2560,128,0.009956444303194681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,128,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,64,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,32,0.01109688894616233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,65536,0.38779465357462567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,12288,0.1401288906733195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,65536,0.2560613420274523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,51200,0.3738995658026801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,16384,0.11836532751719157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2560,768,0.011048888994587792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,12288,0.08998666869269477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2560,1536,0.02352711061636607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,51200,0.20373155011071098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,65536,0.6370346811082628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,10240,0.08090311288833618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,12288,0.04846133457289802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,16384,0.15796266661749944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,12288,0.11774755848778619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,8192,0.06710844569736056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,10240,0.04248088929388258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,7168,0.055083556307686694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,10240,0.09895822074678208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,8192,0.0328800015979343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,6144,0.04735111196835836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,8192,0.0808782246377733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,7168,0.030897776285807293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,5120,0.03834933373663161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,7168,0.0719368921385871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,6144,0.02762311034732395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,4096,0.030192000998391047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,6144,0.061625778675079346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,5120,0.023528888821601868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,3584,0.028199109766218398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,5120,0.05196711089875963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,4096,0.019966221517986722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,3584,0.017895999881956313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,3584,0.03849155704180399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,2560,0.020777778493033517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,3072,0.016561778055297006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,16384,0.06858844227261014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,2048,0.01644888851377699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,2560,0.015188443991872998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,2560,0.029440889755884807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,1536,0.013184888495339287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,2048,0.014163555370436775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,2048,0.024923556380801733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,1024,0.00961599995692571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,1536,0.012208888928095499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,1536,0.02056711084312863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,768,0.007923555870850882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,1024,0.015834665960735746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,3072,0.023200000325838726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,768,0.010304888917340172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,512,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,768,0.01347644461525811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,4096,0.043046222792731396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,256,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,128,0.004495111190610462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,512,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,256,0.009321778184837764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,64,0.0038906666967603895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,512,0.011330666641394297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,2048,32,0.0041884444653987885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,128,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,51200,0.5068391164143881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,64,0.009451555709044138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,32,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,65536,0.3160088857014974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,65536,0.1841191053390503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,1024,0.010796444283591377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,16384,0.09732533163494533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,65536,0.5315235455830892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,51200,0.147544887330797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,16384,0.049710220760769315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,51200,0.41745689180162215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,16384,0.13235910733540854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,256,0.010191111101044549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,10240,0.06156177653206719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,12288,0.10058044724994236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,10240,0.03389599919319153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,8192,0.04639910989337497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,10240,0.08389866352081299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,7168,0.03920000129275852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,8192,0.025567111041810777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,8192,0.06784355640411377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,51200,0.25973065694173175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,7168,0.02368444369898902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,6144,0.03314400050375197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,5120,0.028367999527189467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,7168,0.06031911240683662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,6144,0.020896888441509668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,4096,0.02253155575858222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,6144,0.052096890078650586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,5120,0.018632888793945312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,12288,0.06770133309894137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,3584,0.0211173329088423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,5120,0.04470400015513102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,3072,0.01770666738351186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,4096,0.01608711149957445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,2048,3072,0.034444444709353976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,4096,0.03729866610633002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,3584,0.01498222185505761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,12288,0.03808800048298306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,2560,0.016176000237464905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,2048,0.013032889200581444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,3072,0.013808888693650564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,3584,0.03330933385425144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,1536,0.010592000352011787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,3072,0.029680000411139593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,2048,0.011552888486120435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,1024,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,2560,0.025835555460717943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,2048,0.0217866665787167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,1536,0.010625777973069085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,768,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,1024,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,512,0.005882666756709416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,1024,0.013886222408877479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,768,0.009310222334331935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,256,0.004537777768241035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,768,0.01202133297920227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,512,0.008745777938101027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,128,0.0038408889538711975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,512,0.01016177816523446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,256,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,64,0.0034942221310403612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,256,0.008966222405433655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1536,32,0.0035111109415690103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,128,0.008346666892369589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,64,0.00831822223133511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,2048,128,0.00905777762333552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,32,0.008415111237102086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,65536,0.2565893332163493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,2560,0.012780444489585029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,51200,0.16667732927534315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,65536,0.171888894504971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,16384,0.0664551125632392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1536,1536,0.018086221483018663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,51200,0.13826489448547363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,12288,0.047762665483686656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,16384,0.04709511001904806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,65536,0.3926764329274495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,51200,0.3174640072716607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,16384,0.10300622383753459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,12288,0.03671733207172818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,8192,0.03201866812176175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,10240,0.032864888509114586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,7168,0.02735111117362976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,10240,0.06469333171844482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,8192,0.02472355630662706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,6144,0.023606222536828782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,8192,0.05272355675697327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,7168,0.022100443641344707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,5120,0.0195448895295461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,7168,0.0470746656258901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,4096,0.01649777756796943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,6144,0.04122222132152981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,5120,0.017722666263580322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,3584,0.01463822192615933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,5120,0.03497066762712266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,4096,0.01578044394652049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,4096,0.02903377678659227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,3072,0.013376889129479727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,3584,0.014706666270891825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,2560,0.01088355564408832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,3072,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,10240,0.04178577661514282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,3072,0.023358222511079576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,2048,0.009388444324334463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,2560,0.012776888906955719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,1536,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,12288,0.07747466696633233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,2560,0.020438222421540152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,2048,0.011356444822417365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,1024,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1536,128,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,2048,0.017542221479945712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,1536,0.010392888552612728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,1024,0.009572444690598382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,1536,0.014779556128713818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,512,0.004528000122971005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,1024,0.011719110939237805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,256,0.0036231109665499795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,6144,0.019831111033757527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,768,0.010105777945783403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,512,0.008647110727098253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,128,0.0031715554909573663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,512,0.008795555267069075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,64,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,128,0.008268444074524773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,256,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,32,0.003169777699642711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,128,0.007369777394665613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,64,0.007229333122571309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,65536,0.21421866946750215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,32,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,51200,0.16112889183892146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,65536,0.1700577735900879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,65536,0.3450026777055528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,1024,768,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,51200,0.1369102266099718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,51200,0.27498311466640896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,12288,0.04162933429082235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,768,0.009287111461162567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,16384,0.046279999944898814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,10240,0.03219644559754266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,16384,0.09056355555852254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,12288,0.03705777724583944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,8192,0.02663822140958574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,12288,0.06856977939605713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,1024,256,0.008290666672918532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,10240,0.03264177838961283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,7168,0.02380888826317257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,10240,0.057778669728173145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,8192,0.02392533255947961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,8192,0.04912622107399834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,7168,0.02184533410602146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,1024,3584,0.02626222206486596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,5120,0.017688888642523024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,7168,0.04212888744142321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,6144,0.02055111030737559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,4096,0.015102222561836243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,6144,0.0370142228073544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,5120,0.017445334129863314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,3584,0.013210666676362356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,4096,0.015507555670208402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,5120,0.031714667876561485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,3072,0.01180622229973475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,4096,0.026312000221676294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,16384,0.050796445873048574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,2560,0.010016889207892949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,3584,0.023708444502618577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,2048,0.008630221916569604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,3072,0.021351999706692163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,2560,0.012422222230169507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,1536,0.007363555332024892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,2560,0.01863822175396813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,2048,0.011386666860845355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,2048,0.0159599996275372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,1024,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,1536,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,768,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,1536,0.01344622257683012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,1024,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,512,0.004512000001139111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,6144,0.020283554991086323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,1024,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,768,0.009034666750166152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,256,0.0035075553589397003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,768,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,128,0.003208000005947219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,256,0.00833777752187517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,512,0.008150222400824228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,64,0.0028515555378463534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,256,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,768,32,0.002809777855873108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,128,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,3584,0.014530667000346713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,3072,0.013418667018413544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,64,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,32,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,51200,0.12199110454983181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,65536,0.1257688866721259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,16384,0.04213066564665901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,65536,0.3025413354237874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,51200,0.10175733433829413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,12288,0.030268443955315485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,51200,0.23802844683329263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,16384,0.03567644291453891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,10240,0.023715555667877197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,16384,0.08016977707544963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,12288,0.026755554808510676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,8192,0.01919911139541202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,768,512,0.008636444807052612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,12288,0.060106668207380504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,7168,0.016688888271649677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,10240,0.024263999528355066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,6144,0.015104888214005364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,10240,0.05066577924622429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,8192,0.04167822334501478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,7168,0.016553777787420485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,768,128,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,5120,0.013002666334311167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,7168,0.03728444377581278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,4096,0.010846222440401712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,6144,0.015184000134468079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,3584,0.010011555420027839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,6144,0.032794667614830866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,4096,0.012102222277058495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,5120,0.02788355615403917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,3072,0.009759110709031424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,4096,0.02346399923165639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,3584,0.011428444749779172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,2560,0.00871022211180793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,3584,0.021167111065652635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,3072,0.010946667028797997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,2048,0.007583111524581909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,3072,0.019021333919631112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,2560,0.01036888857682546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,1536,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,2560,0.01685066686736213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,2048,0.009740444521109263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,1024,0.00510044437315729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,2048,0.014475555883513557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,1536,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,768,0.004516444272465176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,1536,0.012208888928095499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,1024,0.00831733312871721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,512,0.003882666842805015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,768,0.00814400033818351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,8192,0.01791200041770935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,768,0.00867733359336853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,512,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,512,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,128,0.002536000063021978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,256,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,65536,0.1478453344768948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,256,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,5120,0.013615999784734515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,32,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,512,128,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,128,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,65536,0.10221866766611735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,64,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,32,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,51200,0.08444533083173965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,65536,0.12540000014834934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,16384,0.03076710965898302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,65536,0.25912711355421275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,51200,0.1004284421602885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,12288,0.021727111604478624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,51200,0.2036986615922716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,512,1024,0.009722666607962714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,16384,0.03496177660094367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,256,0.0031395554542541504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,16384,0.06792444652981229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,12288,0.026570666167471144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,8192,0.013555555707878537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,10240,0.024007999234729346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,12288,0.051755554146236844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,7168,0.012160888976520963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,8192,0.017848889032999676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,512,64,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,10240,0.04365155431959364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,6144,0.010763555765151978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,5120,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,8192,0.03613777624236213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,7168,0.016347555650605094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,6144,0.015038222074508667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,4096,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,7168,0.03239377670817905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,5120,0.013432888521088494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,6144,0.028382221857706707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,5120,0.0244168887535731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,3072,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,3584,0.011549333731333414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,4096,0.020648888415760465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,2560,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,3584,0.01867644488811493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,3072,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,2048,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,2560,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,2560,0.014910222755538093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,1536,0.004232888834344016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,2048,0.009736888938479954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,2048,0.012885332935386233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,1024,0.0034924443397257063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,1536,0.00904088881280687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,768,0.0032133333798911837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,1536,0.011019555230935415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,1024,0.00831288927131229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,512,0.002838222309947014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,1024,0.008990222381220924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,768,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,256,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,768,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,3584,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,128,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,4096,0.012056888805495368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,512,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,256,0.007355555891990662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,64,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,32,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,128,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,256,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,128,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,65536,0.06468444400363497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,3072,0.016921778519948322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,51200,0.05346399876806471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,64,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,256,32,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,16384,0.018274666534529794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,256,10240,0.0185208883550432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,65536,0.1244879961013794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,51200,0.10012088881598578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,65536,0.22125866678025988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,16384,0.034767998589409724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,10240,0.01275466630856196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,16384,0.058317330148484975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,12288,0.02657777733272976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,8192,0.01275111072593265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,12288,0.04453955425156487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,10240,0.02437244521247016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,7168,0.0110880004035102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,10240,0.0377768874168396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,8192,0.018204443984561496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,6144,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,8192,0.030913776821560327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,7168,0.01688888834582435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,256,512,0.007592889169851939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,5120,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,7168,0.02770488957564036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,4096,0.007240888973077138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,6144,0.024493333366182115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,5120,0.013812444276279874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,3584,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,5120,0.02124711043304867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,4096,0.012181333369678922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,4096,0.01794399983353085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,3072,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,3584,0.012000888586044312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,2560,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,3584,0.016541333662139047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,3072,0.011027555498811932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,2048,0.004813333352406819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,3072,0.014917333920796713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,2560,0.010494222243626913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,1536,0.0041866666740841335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,2560,0.013418667018413544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,2048,0.009744000103738572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,1024,0.003273777870668305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,2048,0.011756444142924415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,1536,0.009032888544930352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,768,0.0031679999083280563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,1536,0.010039110978444418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,1024,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,512,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,1024,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,768,0.008005333443482717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,6144,0.015439111325475903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,256,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,768,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,512,0.007601777712504069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,128,0.0021280000607172647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,256,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,64,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,256,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,32,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,128,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,65536,0.06435733371310763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,64,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,128,12288,0.012183999849690331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,51200,0.05041422115431892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,32,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,51200,0.1734213299221463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,12288,0.013612444202105204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,65536,0.21949688593546549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,16384,0.0582853356997172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,51200,0.173088894950019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,8192,0.010568888651000129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,12288,0.04431644413206312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,7168,0.010037333601050908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,6144,0.008585777547624376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,10240,0.037520001331965126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,5120,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,8192,0.030681778987248737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,7168,0.02736888825893402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,4096,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2817,128,128,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,3584,0.006331555545330048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,6144,0.02416177756256527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,5120,0.021120000216695998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,2560,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,4096,0.018067555295096505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,2048,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,3584,0.016376889414257474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,1536,0.003946666502290302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,16384,0.014783999986118741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,1024,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,2560,0.013271110753218332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,768,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,2048,0.01168444421556261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,512,0.0026693333768182327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,1536,0.010000000397364298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,256,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,1024,0.00830488900343577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,128,0.0021395554973019496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,768,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,64,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,512,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,32,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,256,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,65536,0.06394488943947686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,128,512,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,51200,0.04884622163242764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,64,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,3072,0.0058400001790788435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,16384,0.013219555219014486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,32,0.005157333281305102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,12288,0.01035733355416192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,10240,0.009353777600659264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,65536,0.22086221641964385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,3072,0.0148026661740409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,8192,0.00847822262181176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,51200,0.17275555928548178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,7168,0.008228444390826756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,6144,0.007560888926188151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,10240,0.0376160012351142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,5120,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,8192,0.030658665630552504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,64,10240,0.012062222593360476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,7168,0.02737244466940562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,4096,0.008637333081828224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,3584,0.008184000021881526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,6144,0.02421244482199351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,3072,0.007204444044166141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,5120,0.02124888863828447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,2560,0.006254222244024277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,4096,0.01794577803876665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,2048,0.004521777646409141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,3584,0.01646577815214793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,64,128,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,1536,0.003867555409669876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,3072,0.014803555276658801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,1024,0.003220444545149803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,768,0.0028942223224374983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,2048,0.011518222590287527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,1536,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,512,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,1024,0.008282666405042013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,256,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,128,0.0021137777302000257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,768,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,64,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,12288,0.04434222314092848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,16384,0.0581173300743103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2817,32,32,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,512,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,256,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,128,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,64,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,32,0.005212444398138259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2817,32,2560,0.013386666774749756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,16384,1.1591911315917969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,16384,3.9171795315212674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,16384,3.928594801161024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,51200,4.256551954481337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,16384,3.386419508192274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,12288,2.9269057379828562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,12288,0.8482391039530436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,12288,2.827061335245768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,12288,2.550142288208008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,10240,0.7051288816663953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,10240,2.580049726698134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,10240,2.272360059950087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,8192,0.5762524604797363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,8192,1.9341085222032335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,10240,2.1509511735704208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,8192,1.969455083211263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,7168,0.4974168671502008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,8192,1.6607359780205622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,51200,11.466072930230034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,7168,1.8748533460828993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,7168,1.5741484959920247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,6144,0.4427475664350722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,7168,1.4940889146592882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,6144,1.3462595409817164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,6144,1.6333689159817164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,5120,1.1786648432413738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,5120,0.3662239975399441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,5120,1.1431173748440213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,6144,1.143872896830241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,4096,1.020127084520128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,4096,0.3158693313598633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,5120,1.029205322265625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,4096,0.8775324291653104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,3584,0.8230169084337023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,4096,0.7758577664693197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,3584,0.27803288565741646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,3584,0.8046284251742892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,3072,0.6322186787923177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,3072,0.24604890081617567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,3584,0.7424471113416883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,3072,0.6978649033440484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,2560,0.5930302408006456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,2560,0.2248142295413547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,2560,0.5680204497443305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,3072,0.6174542109171549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,2048,0.4143599934048123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,2560,0.5198062260945638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,2048,0.1837902201546563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,1536,0.3790817790561252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,2048,0.49347199334038633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,1536,0.38204267289903426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,1536,0.1479359997643365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,1024,0.2605031066470676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,2048,0.4370604356129964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,51200,11.557973225911459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,768,0.2025057739681668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,1024,0.11644000477261013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,51200,16.956819322374134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,1024,0.2737270991007487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,1536,0.338447994656033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,512,0.15019910865359837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,1024,0.24898576736450195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,768,0.11486844221750896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,768,0.2276862197452121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,256,0.09881066613727146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,512,0.09194311168458726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,512,0.19472889105478922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,128,0.07664089070426093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,256,0.08243555492824979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,512,0.16677244504292807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,256,0.1656284464730157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,256,0.16005155775282118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,65536,128,0.07784355349010892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,32,0.0689528915617201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,128,0.1568275557623969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,128,0.10246577527787949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,32,0.160680890083313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,65536,768,0.2883973386552599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,65536,64,0.06753600305981107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,65536,64,0.15868443912929958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,51200,3.1916222042507596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,65536,4.226288689507379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,16384,3.35932985941569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,51200,8.538790384928385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,16384,3.294334199693468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,65536,10.724840799967447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,16384,0.8344160185919868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,12288,2.229435602823893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,12288,0.6510791248745388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,16384,2.666960822211372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,12288,2.2063253190782333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,12288,1.9640551673041449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,51200,9.299550374348959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,51200,13.781541612413195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,10240,2.121198230319553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,10240,0.5411004490322536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,10240,1.8301785786946614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,8192,0.4361262321472168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,10240,1.6768995920817058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,8192,1.4338595072428386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,8192,1.457980473836263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,7168,1.1812755796644423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,7168,0.4181528886159261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,65536,17.093236287434895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,8192,1.2923360400729709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,7168,1.219029320610894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,65536,11.889887491861979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,6144,1.0866720411512587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,6144,0.3452133337656657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,7168,1.1405902438693576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,6144,1.0537688997056749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,5120,0.3056524329715305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,5120,0.991554684109158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,4096,0.6939848793877496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,5120,0.849131531185574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,6144,0.9920995500352648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,4096,0.24612087673611113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,4096,0.6918409135606555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,5120,0.762247986263699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,3584,0.22891110844082305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,3584,0.746854199303521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,3072,0.5614524417453343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,3584,0.6547004381815592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,3584,0.552079995473226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,3072,0.19519555568695068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,2560,0.43023912111918133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,2560,0.16860000292460123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,3072,0.4781182077195909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,2048,0.3487857712639703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,2560,0.6171458032396104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,2048,0.14185244507259792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,2560,0.4100026554531521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,1536,0.28383199373881024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,2048,0.39190488391452366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,4096,0.6241048706902398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,1024,0.19482844405704072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,1536,0.11618755923377143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,2048,0.3368142180972629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,1536,0.29225156042310924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,768,0.14663378397623697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,1024,0.1057164470354716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,1024,0.21592088540395102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,1536,0.26698043611314565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,3072,0.5168631341722276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,512,0.1086497770415412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,768,0.0815111133787367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,768,0.17715377277798125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,1024,0.19803288247850206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,256,0.07358755668004353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,512,0.07398044400744967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,512,0.14613510502709284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,128,0.06655555301242404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,512,0.13497955269283718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,256,0.06512533293830024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,768,0.16493154896630183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,64,0.051466666989856295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,51200,32,0.052904887331856616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,51200,128,0.06489066945181952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,256,0.10556177960501777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,256,0.13099111451043025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,51200,128,0.08033955759472318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,64,0.1253582239151001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,32,0.1273582246568468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,51200,128,0.12381066216362847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,65536,1.2854933208889432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,51200,1.0127795537312825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,16384,1.0895261764526367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,51200,2.9937235514322915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,65536,3.693090650770399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,16384,0.2937733332316081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,16384,1.136779573228624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,51200,4.132856157090929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,12288,0.748569753434923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,16384,0.8634666866726346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,51200,2.9523991478814016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,12288,0.2533173296186659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,12288,0.7074862056308322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,65536,5.193773481580946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,10240,0.5421804322136773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,12288,0.6153590944078233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,10240,0.1813431051042345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,65536,3.7881893581814237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,8192,0.4560737609863281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,10240,0.6150995360480415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,8192,0.14748088518778482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,8192,0.48854223887125653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,10240,0.520402643415663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,7168,0.41302045186360675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,7168,0.1321857770284017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,8192,0.41272711753845215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,6144,0.2940133412679036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,6144,0.11594755119747585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,7168,0.35695823033650714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,5120,0.2666400008731418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,6144,0.3571351104312473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,7168,0.4626426696777344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,5120,0.09959555334515041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,4096,0.19153243965572783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,5120,0.2863955497741699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,6144,0.32179734441969127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,3584,0.1925039953655667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,4096,0.08603288729985555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,5120,0.2716391086578369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,3584,0.07723377810584174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,4096,0.25200088818868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,3072,0.15100799666510686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,4096,0.21566666497124565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,3584,0.19997066921657988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,2560,0.12922222084469265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,3072,0.06728621986177233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,3584,0.1901093324025472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,3072,0.1738711065716214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,2560,0.05903999672995674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,3072,0.16784444120195177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,2560,0.17511467138926187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,1536,0.07567644119262695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,2048,0.051515556044048734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,2560,0.14288355244530573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,2048,0.12905332777235243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,1024,0.06404978036880493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,2048,0.12029777632819282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,1536,0.042224887344572276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,1536,0.09807555543051825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,768,0.05116622315512764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,1024,0.03530044356981913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,1536,0.09943466716342503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,512,0.038337777058283486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,768,0.03251466817326016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,1024,0.07657244470384386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,768,0.06059110826916165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,256,0.024263999528355066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,768,0.06365778048833211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,512,0.0518817769156562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,128,0.022096888886557683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,512,0.05303288830651177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,256,0.02713777787155575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,256,0.042136000262366406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,64,0.023511111736297607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,128,0.026120889517996047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,128,0.04393511017163595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,32,0.024639111426141527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,16384,128,0.03194844391610887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,64,0.044293334086736046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,32,0.04478222131729126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,1024,0.07520800166659884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,16384,512,0.029828445778952703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,16384,256,0.046201778782738574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,16384,2048,0.09886044263839722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,65536,0.9917457368638781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,51200,0.7762275801764594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,16384,0.8541288905673556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,65536,2.784975899590386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,51200,2.3477937910291886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,16384,0.22233867645263672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,16384,0.8135084576076932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,12288,0.5802008840772841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,16384,0.630308469136556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,65536,3.8444959852430554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,12288,0.5381795565287272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,51200,3.453066719902886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,10240,0.4893031120300293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,12288,0.47472180260552305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,51200,2.253246307373047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,65536,2.9239732954237194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,10240,0.15518044100867376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,10240,0.455185784233941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,8192,0.41466310289171004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,10240,0.39420000712076825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,8192,0.1268275578816732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,8192,0.35125067498948837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,7168,0.34855021370781797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,8192,0.3404720094468858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,6144,0.27679377131991917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,12288,0.18292800585428873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,7168,0.3228257762061225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,6144,0.09035822418000962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,5120,0.24163466029696992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,6144,0.2952968809339735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,6144,0.2445440027448866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,4096,0.1687697834438748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,5120,0.07892977529101901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,5120,0.2235422134399414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,5120,0.2064088847902086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,4096,0.0655297769440545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,3584,0.14347378412882486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,4096,0.17734666665395102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,4096,0.17571466498904756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,3072,0.11537955866919623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,3584,0.05785777833726671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,7168,0.10574933555391099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,3584,0.15380711025661892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,7168,0.30296889940897626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,3584,0.14909511142306858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,3072,0.05199911197026571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,2048,0.08536977900399102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,3072,0.14160533746083578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,2560,0.04661422305636936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,3072,0.13235199451446533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,2560,0.1155351135465834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,1536,0.060200889905293785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,2048,0.040264000495274864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,2560,0.11342488394843207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,2048,0.0957875582906935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,1024,0.051651553975211255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,2048,0.09654400083753799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,1536,0.033434665865368314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,1536,0.07591022385491265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,768,0.04000533289379544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,1024,0.0280693338976966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,1536,0.07964799801508586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,1024,0.05853333075841268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,512,0.0266942216290368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,1024,0.060976001951429576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,768,0.025895110434956018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,768,0.04659466611014473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,512,0.023803555303149756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,768,0.05161777800983853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,128,0.015207111835479736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,512,0.03897600041495429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,2560,0.11092978053622776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,512,0.043755554490619235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,256,0.020962667134073045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,64,0.020861332615216572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,256,0.0343795551194085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,12288,128,0.020257777637905545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,128,0.03277600142690871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,12288,128,0.025788444611761305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,64,0.0329751107427809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,32,0.03334577878316244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,256,0.02660177813635932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,12288,256,0.034865776697794594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,12288,32,0.021365332934591506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,65536,0.7807964218987359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,51200,1.6487627029418945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,16384,0.6952995724148221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,65536,2.1222142113579645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,51200,0.6234729025099013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,16384,0.1866968870162964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,12288,0.5133848720126682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,16384,0.6419857872856988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,16384,0.5290373166402181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,12288,0.14201777511172825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,12288,0.4514017634921604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,51200,2.536599900987413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,10240,0.34354133076137966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,65536,3.4158905877007375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,65536,2.4361644321017795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,12288,0.44187198744879824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,10240,0.12283644411298965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,8192,0.3078960047827827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,10240,0.3711831039852566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,8192,0.09837866491741604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,10240,0.32780443297492134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,7168,0.24231200748019746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,8192,0.3094782299465603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,7168,0.08799555566575791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,6144,0.21973866886562773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,8192,0.3264862166510688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,7168,0.2947057882944743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,6144,0.07741600275039673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,5120,0.18789777490827772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,6144,0.22962400648328993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,6144,0.20389778084225127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,4096,0.1519600020514594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,5120,0.18943377335866293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,3584,0.11227822303771973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,5120,0.17377244101630318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,4096,0.14829244878556994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,4096,0.14209689034356013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,3584,0.050844444168938525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,3584,0.13021688991122776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,3072,0.11389777395460342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,3584,0.12964267200893825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,3072,0.044863111442989774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,3072,0.1140506664911906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,7168,0.23274132940504286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,2560,0.0833395587073432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,3072,0.11169421672821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,2048,0.07306044631534152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,5120,0.06683733065923055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,2560,0.040256887674331665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,2560,0.0986257791519165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,1536,0.05085688829421997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,2048,0.034881777233547635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,4096,0.06570044491026136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,2048,0.08145599895053439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,1024,0.03603022297223409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,1536,0.028696888022952612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,51200,1.9146675533718531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,1536,0.06520889202753703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,768,0.027085334062576294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,1536,0.0686968896124098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,1024,0.024079999989933435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,1024,0.04939644535382589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,512,0.020038222273190815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,1024,0.05263377891646492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,768,0.040052443742752075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,768,0.021993777818149988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,768,0.0454311105940077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,256,0.01553600033124288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,512,0.01955911186006334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,512,0.03241511185963949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,128,0.014410666293568082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,512,0.03760533200369941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,256,0.01759999990463257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,256,0.028718223174413044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,64,0.012431110772821637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,256,0.03004266818364461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,2560,0.09754400120841132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,10240,32,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,10240,128,0.0181013329161538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,128,0.027638221780459087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,128,0.023405333360036213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,10240,2048,0.08420799838172065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,64,0.02781244450145298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,10240,32,0.027320888307359483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,51200,0.523813353644477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,65536,0.6780257754855685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,51200,1.5024755265977647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,16384,0.5676657888624403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,16384,0.6994400024414062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,16384,0.16641688346862793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,65536,2.183349397447374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,12288,0.3599048985375299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,51200,2.2437110477023654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,51200,1.5558844672309027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,16384,0.4512649112277561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,12288,0.40999555587768555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,12288,0.11930755774180095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,10240,0.30176443523830837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,12288,0.3385768996344672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,10240,0.10137777858310276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,8192,0.2249128818511963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,10240,0.33652533425225156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,10240,0.2787422339121501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,8192,0.08191821972529094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,7168,0.18720000320010713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,8192,0.26155021455552846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,8192,0.22740443547566733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,7168,0.07341066996256511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,6144,0.16174488597446018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,7168,0.21607466538747153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,5120,0.13641599814097086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,6144,0.06461066669887967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,7168,0.1997671127319336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,6144,0.18506488535139295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,4096,0.11342667208777533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,5120,0.05587999688254463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,6144,0.17295644018385145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,5120,0.15252355734507242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,3584,0.10152533319261338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,4096,0.0462719996770223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,5120,0.1565635601679484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,4096,0.12274666627248128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,3072,0.08674489127265082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,65536,2.076441870795356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,3584,0.04169955518510607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,4096,0.12498222457038032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,3584,0.10775022374259101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,2560,0.06852889060974121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,3584,0.10945510864257812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,3072,0.09504799710379706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,2048,0.06051199966006809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,2560,0.033597333563698664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,3072,0.09803911050160725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,1536,0.043841779232025146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,2048,0.028804444604449805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,2048,0.06781511174307929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,2048,0.07070400317509969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,1536,0.024098666177855596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,1024,0.028571556011835735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,1536,0.058485335773891874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,65536,2.7714150746663413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,1024,0.020670221911536325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,1024,0.041469332244661115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,768,0.022983999715911016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,1024,0.04604622059398227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,3072,0.03736622134844462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,768,0.018537777993414137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,512,0.017495110630989075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,768,0.0340008901225196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,768,0.039493332306543984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,2560,0.08210488822725084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,256,0.013154666456911298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,512,0.01648355523745219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,2560,0.08492177724838257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,512,0.0273324449857076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,128,0.013626666532622443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,512,0.033040000332726374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,256,0.015307555596033732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,64,0.012689777546458773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,256,0.023592889308929443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,256,0.02604622145493825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,8192,32,0.01278133359220293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,8192,128,0.015791111522250705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,128,0.023024888502226934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,8192,128,0.01993955506218804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,64,0.023009777069091797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,32,0.02289955483542548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,65536,0.6215102407667372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,8192,1536,0.05355822377734714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,65536,1.5581955379909938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,51200,1.3955661985609267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,51200,0.4985040028889974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,51200,1.8754196166992188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,65536,2.263334274291992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,16384,0.15339377191331652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,16384,0.40323911772833926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,16384,0.6021048757765028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,12288,0.30638933181762695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,65536,1.7162649366590712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,12288,0.11180267069074844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,10240,0.2903582255045573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,12288,0.33085065417819554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,10240,0.0925253364774916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,12288,0.3087582323286268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,8192,0.23969954914516875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,10240,0.2977608839670817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,51200,1.3791946834988063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,10240,0.27587734328375924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,8192,0.07452711131837633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,16384,0.4076968828837077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,8192,0.21217332945929634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,7168,0.06880088647206624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,8192,0.21192444695366752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,6144,0.18432266182369658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,7168,0.1901102198494805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,5120,0.1492275529437595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,6144,0.05829422341452705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,6144,0.1597813367843628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,7168,0.2212817801369561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,4096,0.11350133683946396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,6144,0.1587617794672648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,5120,0.054655111498302884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,5120,0.1361066632800632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,5120,0.13564088609483507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,3584,0.0938168896569146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,4096,0.04217333263821072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,4096,0.10816444291008843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,3072,0.08110222551557753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,4096,0.11348888609144424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,3584,0.037682665718926325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,3584,0.10161955489052667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,2560,0.07240088780721028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,3584,0.09957155254152085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,3072,0.0846995578871833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,7168,0.20722844865587023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,3072,0.03392444385422601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,2560,0.030788444810443457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,2560,0.07325155867470635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,3072,0.08888355228636001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,1536,0.04042044281959534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,2560,0.07804799742168851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,1024,0.027282666828897264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,1536,0.02176800039079454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,2048,0.0655991103914049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,1536,0.05343911051750183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,768,0.021935999393463135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,1024,0.03677511215209961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,1024,0.01828266680240631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,1024,0.04283466604020861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,512,0.01641155613793267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,768,0.016599999533759225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,768,0.030365334616767034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,768,0.03601510988341437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,256,0.012746666868527731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,512,0.015456888410780164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,512,0.024615110622511968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,512,0.030658665630552504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,256,0.014315555493036905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,2048,0.06000355879465739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,128,0.009876444107956356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,256,0.021402666966120403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,256,0.023706666297382776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,2048,0.02611733310752445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,64,0.008672888908121321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,7168,128,0.014256000518798828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,2048,0.060447997517055936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,7168,32,0.009153777526484596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,128,0.020615999897321064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,1536,0.04847288793987698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,7168,128,0.018350222044520907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,64,0.020661332541041903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,7168,32,0.020619556307792664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,65536,0.5237360000610352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,51200,0.9484657711452908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,65536,1.2822453180948894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,51200,0.40980890062120223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,16384,0.3493377897474501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,51200,1.4818240271674261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,16384,0.12194044060177273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,16384,0.39372976620992023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,51200,1.2533751593695748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,12288,0.26080534193250865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,16384,0.38025689125061035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,65536,1.5954622692532008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,12288,0.28227734565734863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,10240,0.23916710747612846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,12288,0.2779955599043104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,8192,0.1688693364461263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,10240,0.08940266900592381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,10240,0.24894044134351942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,10240,0.23177867465549043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,7168,0.17231467035081652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,8192,0.06487466891606648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,8192,0.26307998763190377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,6144,0.13240977128346762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,7168,0.06199377775192261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,7168,0.16548177931043836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,5120,0.10116710927751328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,65536,1.869970745510525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,6144,0.05251555641492208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,7168,0.16538933912913004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,6144,0.1524088912540012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,4096,0.09278666973114014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,5120,0.04531822270817227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,6144,0.1450328826904297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,12288,0.0941271119647556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,5120,0.11861688560909694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,4096,0.037852442926830716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,5120,0.1229777733484904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,3072,0.0693884425693088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,4096,0.10253244638442993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,3584,0.03421777817938063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,3584,0.08483199940787421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,3072,0.030777778890397813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,2560,0.054008887873755566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,3584,0.09114844269222683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,3072,0.07389244768354628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,2048,0.04366222355100843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,3072,0.08188711272345649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,2560,0.0273617770936754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,8192,0.18743644820319283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,1536,0.03633244501219855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,2560,0.07081688774956597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,2048,0.023747555083698694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,2048,0.053137779235839844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,1024,0.025045333637131587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,1536,0.0203128887547387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,1536,0.04266222318013509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,3584,0.08227377467685275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,1536,0.04929066697756449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,1024,0.017478222648302715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,4096,0.09700444671842788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,1024,0.032489776611328125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,512,0.015028445257080926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,1024,0.03895644346872965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,768,0.01576622161600325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,768,0.027292443646325007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,256,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,768,0.03230133321550157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,512,0.014418666561444601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,512,0.021783111823929682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,512,0.02700622214211358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,256,0.013423110875818463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,256,0.01921955578856998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,2560,0.06411911381615533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,256,0.02185955478085412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,64,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,32,0.008639111287064022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,128,0.0185208883550432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,128,0.017237333787812125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,6144,2048,0.06017866399553087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,64,0.018579555882347953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,6144,32,0.018390221728218924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,768,0.021086222595638696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,65536,0.45914845996432835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,6144,128,0.00963199966483646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,6144,128,0.013813333378897773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,51200,0.9304666519165039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,51200,0.3679857783847385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,16384,0.3157404528723823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,51200,1.2592008378770616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,65536,1.2798560460408528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,16384,0.10536089208390977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,16384,0.3664826552073161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,12288,0.2816630999247233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,16384,0.2994026607937283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,51200,0.9410515891181098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,10240,0.18806666798061797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,65536,1.2594150967068143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,12288,0.08724621931711833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,12288,0.22784890068901908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,8192,0.14621155791812473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,10240,0.07085333267847697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,10240,0.23764178487989637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,8192,0.056983113288879395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,7168,0.1258266634411282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,8192,0.15825955073038736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,65536,1.816296895345052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,6144,0.09574755695131089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,8192,0.15971199671427408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,7168,0.0500239994790819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,5120,0.09107910924487644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,6144,0.0440924432542589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,7168,0.1361804405848185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,6144,0.1210435496436225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,5120,0.037603557109832764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,4096,0.07308355304929945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,5120,0.10199111037784153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,12288,0.2382862302992079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,5120,0.1034088863266839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,3584,0.05801333321465386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,4096,0.03221599923239814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,4096,0.08171644475724962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,4096,0.08513955275217693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,3584,0.029156443145540025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,3072,0.05330933464898003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,3584,0.07326844665739271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,10240,0.19666666454739043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,2560,0.04176355401674906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,3584,0.07846311065885755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,3072,0.0258577780591117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,2048,0.0362426671716902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,3072,0.0638515551884969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,2560,0.023935111032591924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,3072,0.06946844524807401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,2560,0.055458668205473155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,1536,0.028816001282797918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,2560,0.05971022446950277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,2048,0.020862221717834473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,1024,0.01925244430700938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,2048,0.050716443194283374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,1536,0.017904000149832833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,6144,0.11955466535356309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,1536,0.036990222003724836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,768,0.015603555573357476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,1536,0.04315822323163351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,1024,0.015182221929232279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,1024,0.028903110159768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,512,0.01185155577129788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,768,0.01404622197151184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,1024,0.03284622232119242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,256,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,768,0.02790488799413045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,512,0.012841777669058906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,512,0.023896889554129705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,128,0.007330666813585494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,256,0.01200711147652732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,256,0.016879111528396606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,256,0.020187555087937247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,64,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,5120,128,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,7168,0.1450586716334025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,128,0.016399110356966656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,5120,128,0.016480889585283067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,2048,0.046358221107059054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,64,0.016188444362746347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,32,0.016213332613309223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,768,0.023918223049905565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,5120,512,0.019150222341219585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,65536,0.4055448902977838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,65536,1.1186427010430229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,5120,32,0.006670222100284364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,51200,0.3241688940260145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,16384,0.30392177899678546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,65536,1.3496675491333008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,16384,0.10435822274949814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,65536,1.1177244186401367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,16384,0.266208012898763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,12288,0.2007208930121528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,51200,1.0070311228434246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,10240,0.1709351142247518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,16384,0.2884417904747857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,51200,0.7994186613294813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,12288,0.216630220413208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,8192,0.1355431079864502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,10240,0.06212355693181356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,12288,0.21349688371022543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,51200,0.7835395601060656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,8192,0.04931288957595825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,10240,0.1780524386299981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,7168,0.11972888310750325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,8192,0.13258222738901773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,8192,0.14925244119432238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,6144,0.09845599863264297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,7168,0.04393777913517422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,7168,0.11790577570597331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,6144,0.03849066628350152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,7168,0.1244000064002143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,5120,0.07867822382185194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,6144,0.10353866550657485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,4096,0.05698933204015096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,6144,0.10609866513146295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,5120,0.03258488906754388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,12288,0.07304444577958849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,5120,0.08550755845175849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,5120,0.09126755264070298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,3584,0.05452533231841194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,4096,0.02756799923049079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,4096,0.06990133391486274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,3072,0.04574666751755608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,4096,0.0760657787322998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,3584,0.025106666816605464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,3584,0.06261689133114286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,2560,0.037949333588282265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,3072,0.022689777943823073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,3584,0.06844888793097602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,3072,0.05493866735034519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,2048,0.031067556805080835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,3072,0.05968177980846829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,2560,0.02060622142420875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,2560,0.04756711257828606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,1536,0.023434667123688593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,2560,0.05184266633457608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,2048,0.01776444415251414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,2048,0.040017777019076876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,1024,0.017277333471510146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,2048,0.04563200142648485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,1536,0.01588622232278188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,1536,0.032122668292787336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,1536,0.03732444511519538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,1024,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,1024,0.02865333358446757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,512,0.010654222634103564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,768,0.020850666695170935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,768,0.012755555411179861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,10240,0.1688213348388672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,256,0.00793599999613232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,512,0.011694221860832639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,512,0.0166284441947937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,128,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,512,0.02128711177243127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,256,0.014543111125628153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,256,0.011040888726711273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,64,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,256,0.018600889378123812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,32,0.006425777657164468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,128,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,128,0.015207111835479736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,64,0.014095111025704278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,4096,768,0.013372444444232516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,32,0.01385600037044949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,4096,1024,0.024547555380397375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,4096,768,0.024911110599835713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,65536,0.35326578882005477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,65536,0.7680533197191027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,51200,0.7024817996554904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,16384,0.23487644725375703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,4096,128,0.0110871113008923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,51200,0.277200010087755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,65536,1.1202489005194771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,51200,0.8724497689141167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,16384,0.08522133032480876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,12288,0.181621339586046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,16384,0.29116977585686576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,10240,0.14267822106679282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,12288,0.06571733289294772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,16384,0.2764097849527995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,12288,0.2037608888414171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,51200,0.7786951065063477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,8192,0.10625866386625503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,10240,0.05680444505479601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,12288,0.19894399907853869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,7168,0.10080177916420831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,8192,0.04520444406403435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,8192,0.12347733974456787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,10240,0.19516800509558785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,6144,0.08213511440489027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,7168,0.040592888991038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,7168,0.10683733224868774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,5120,0.06450400087568495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,7168,0.11821688546074761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,6144,0.035233779086007014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,6144,0.09487111038631862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,4096,0.052743110391828746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,6144,0.10455644130706787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,5120,0.030631999174753826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,5120,0.07736711369620429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,65536,1.0233768886990018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,4096,0.02628266645802392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,4096,0.06288355588912964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,4096,0.07337688737445407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,10240,0.15330221917894152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,3072,0.03983199927541945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,3584,0.05563822057512072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,3584,0.024108444650967915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,3584,0.06663111183378431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,2560,0.03312266535229153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,3072,0.021916444102923077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,8192,0.14123022556304932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,3072,0.049448887507120766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,2048,0.025787555509143408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,3072,0.058824890189700656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,2560,0.019825778073734708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,2560,0.04276622335116068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,2560,0.05127200153138903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,2048,0.017234666479958426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,2048,0.035903112755881414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,2048,0.04380888740221659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,1024,0.014567999376191033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,5120,0.08765600124994914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,1536,0.01479466590616438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,1536,0.02867822183503045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,3584,0.04667466547754076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,768,0.01201866649919086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,1536,0.03647999962170919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,1024,0.013012444807423485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,512,0.009477333062224919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,1024,0.02203911046187083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,768,0.01237511138121287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,1024,0.028537776735093858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,768,0.018764444523387484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,256,0.007112888826264276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,768,0.024417777856190998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,512,0.011619555453459421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,512,0.01517511076397366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,128,0.006089777582221561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,256,0.01092888911565145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,512,0.020672889219390023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,64,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,256,0.013459555804729462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,32,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3584,128,0.010802666346232096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,256,0.017862222260899015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,128,0.012816888590653738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3584,128,0.0140346669488483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3584,1536,0.020997333857748244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,64,0.01275466630856196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3584,32,0.012898666991127862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,65536,0.29538488388061523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,65536,0.6527466773986816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,51200,0.24541955524020723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,51200,0.5458035469055176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,16384,0.16748089260525176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,12288,0.1353457768758138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,16384,0.08401688602235581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,16384,0.23650132285224068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,51200,0.7771377563476562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,65536,0.984456910027398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,16384,0.22124000390370688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,10240,0.12517600589328343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,65536,0.8783404562208387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,12288,0.05927466683917575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,12288,0.16574754979875353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,8192,0.09198044406043158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,12288,0.17653244071536592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,10240,0.051072001457214355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,10240,0.1350675556394789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,8192,0.04199555516242981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,10240,0.14175555441114637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,8192,0.11161066426171197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,6144,0.06838755475150214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,7168,0.0362737774848938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,7168,0.09684799777136908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,5120,0.061193777455223926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,7168,0.10450844632254706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,6144,0.08149066898557875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,6144,0.08978577454884847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,5120,0.027216888136333887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,5120,0.06887644529342651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,4096,0.04730400111940172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,5120,0.07835200097825792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,51200,0.6787626478407117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,4096,0.023398222194777593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,4096,0.056686222553253174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,4096,0.0648568868637085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,3072,0.034485333495669894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,3584,0.02149155570401086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,7168,0.08210666974385579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,3584,0.05044622222582499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,2560,0.028121779362360638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,3584,0.05835644404093424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,3072,0.019660444723235238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,8192,0.12364266978369819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,3072,0.04472800095876058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,3072,0.05235822333229912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,2560,0.017662222186724346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,2560,0.03895111216439141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,6144,0.03189599845144484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,2560,0.04518577787611219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,2048,0.015760888655980427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,2048,0.032504889700147845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,2048,0.038968000147077776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,1536,0.014133334159851074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,1024,0.014688889185587565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,3584,0.03869333200984531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,1536,0.0329146683216095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,768,0.01204711116022534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,1024,0.012279999752839407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,1024,0.019882667395803664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,1024,0.025401777691311304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,512,0.00903733323017756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,768,0.01717333330048455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,768,0.022283555732833013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,2048,0.02458133300145467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,512,0.014130666851997375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,256,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,512,0.018927110566033255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,1536,0.019290667441156175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,128,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,256,0.01240799989965227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,256,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,256,0.01588888963063558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,64,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,128,0.010069333016872406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,1536,0.02643288837538825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,128,0.01181688904762268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,3072,32,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,3072,128,0.013561777770519257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,64,0.011735999749766456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,3072,32,0.011720000041855706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,768,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,65536,0.5756782425774468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,65536,0.27233955595228404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,51200,0.45165687137179905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,3072,512,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,16384,0.16629689269595677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,51200,0.21606222788492838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,16384,0.06960977448357476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,12288,0.11720977889166938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,16384,0.18155999978383383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,51200,0.6119804382324219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,16384,0.186429328388638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,10240,0.10915288660261367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,12288,0.05567466550403171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,51200,0.5678035418192545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,12288,0.13617422845628527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,8192,0.08936622407701279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,12288,0.14428977171579996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,10240,0.11471466223398845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,10240,0.12449067168765598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,7168,0.06860266791449653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,8192,0.09280444516075982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,8192,0.1014488869243198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,7168,0.03349155518743727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,65536,0.7736124462551541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,6144,0.06314311424891154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,7168,0.0828275548087226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,7168,0.08908177746666802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,5120,0.051670223474502563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,6144,0.03054222133424547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,65536,0.7251173125372993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,6144,0.07135111093521118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,4096,0.039191110266579524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,6144,0.0786879989835951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,5120,0.02685866587691837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,5120,0.060164444976382785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,3584,0.0328764451874627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,5120,0.06655466556549072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,4096,0.02202755543920729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,4096,0.048172444105148315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,3072,0.03052799900372823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,10240,0.04832888974083794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,4096,0.05593422386381361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,3584,0.020449777444203694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,3584,0.04360266526540121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,2560,0.02366133365366194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,3584,0.05079288946257698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,3072,0.018546667363908555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,2048,0.01960088809331258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,2560,0.017307556337780423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,3072,0.03844088978237576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,3072,0.04576444294717577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,2560,0.033462223079469465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,1536,0.016395555602179635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,2048,0.015419556034935845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,2560,0.04019644525316026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,1024,0.01221333361334271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,1536,0.012725333372751871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,2048,0.033879998657438494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,768,0.009679999616410997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,1536,0.02314400010638767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,1536,0.028344889481862385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,1024,0.011186666786670685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,512,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,1024,0.01756266587310367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,1024,0.022355554832352534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,768,0.010704889065689512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,768,0.01516533394654592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,256,0.005534222142563925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,768,0.01962488889694214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,512,0.012434666355450949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,512,0.018241778016090393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,128,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,256,0.009509333305888707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,256,0.011049778097205691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,256,0.01517599986659156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,64,0.004514666481150521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,128,0.009648000200589498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,128,0.010694222317801582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2560,32,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2560,128,0.0123848890264829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,2048,0.02811466654141744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,64,0.010401777923107147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2560,32,0.01074577785200543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,65536,0.4579333199395074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,65536,0.23620266384548613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,51200,0.34003199471367734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,512,0.00998222248421775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,16384,0.12890311082204184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,65536,0.6356168852912055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,51200,0.18702310985989043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2560,8192,0.037159111764695905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,51200,0.5210302140977647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,65536,0.7057778040568033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,16384,0.062011553181542285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,16384,0.15653867191738552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,16384,0.186063117451138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,10240,0.08137777778837416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,12288,0.047672000196244985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,12288,0.12234755357106526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,12288,0.1434239943822225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,10240,0.0982924434873793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,10240,0.1195404397116767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,7168,0.05638933181762695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,8192,0.03154222170511881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,8192,0.08069599999321832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,8192,0.09767111142476399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,7168,0.02860800094074673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,7168,0.07116355498631795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,6144,0.04763022065162659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,7168,0.08629155821270412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,12288,0.09744444158342148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,5120,0.04061333338419596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,6144,0.060588445928361684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,6144,0.07581955856747098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,4096,0.030559112628300984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,5120,0.022244445151752893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,5120,0.051537778642442494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,10240,0.04117688867780898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,5120,0.0649991101688809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,3584,0.027543110979927912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,4096,0.01937244501378801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,4096,0.04270755582385593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,3072,0.02310577697224087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,4096,0.05459200011359321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,3584,0.01794755624400245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,3584,0.038350221183564924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,2560,0.021912000245518152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,3584,0.049588445160124034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,3072,0.01644799941115909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,3072,0.0340924428568946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,2048,0.01798311041461097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,2560,0.015156444576051501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,2560,0.02936444348759121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,2560,0.038254221280415855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,6144,0.025524444050259058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,1536,0.012764444781674279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,2048,0.014064888159434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,2048,0.02497688929239909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,1024,0.009445333646403419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,51200,0.5513866742451986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,2048,0.03312089045842489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,1536,0.012452444268597497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,768,0.00774133370982276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,1536,0.02050755586889055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,8192,0.0676399999194675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,1536,0.027307555079460144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,1024,0.016234666109085083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,512,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,1024,0.021718222233984206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,768,0.01033155537313885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,768,0.013448889056841532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,256,0.004986666556861666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,768,0.01891644464598762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,512,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,512,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,128,0.0042444442709287005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,512,0.017893332574102614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,256,0.009127111070685917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,256,0.009858667022652095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,256,0.014984889162911309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,128,0.009138666921191746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,32,0.0041599998043643106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,3072,0.04395022326045566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,128,0.009406222237481011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,2048,128,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,64,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,2048,32,0.009493333597977957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,65536,0.3177119890848796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,65536,0.18094933032989502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,51200,0.23873689439561632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,2048,1024,0.010775110787815519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,16384,0.09273955557081436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,65536,0.5257626639472114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,65536,0.562155564626058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,51200,0.14374311765034994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,51200,0.4285564422607422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,12288,0.07042488786909315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,16384,0.0484426650736067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,16384,0.13122844696044922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,2048,64,0.003942222230964237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,51200,0.43989067607455784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,16384,0.14962132771809897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,12288,0.039099554220835366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,12288,0.10172888967725967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,8192,0.048342221313052706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,10240,0.03400444322162204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,12288,0.11573155721028645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,10240,0.08307733138402303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,7168,0.04171466827392578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,8192,0.025410667061805725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,8192,0.06753244664933947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,6144,0.0354639987150828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,8192,0.08055555820465088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,7168,0.023193778263198003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,7168,0.05956355730692545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,5120,0.03045066528850132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,7168,0.07144622008005778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,6144,0.02087644404835171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,4096,0.024008888337347243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,6144,0.0521360006597307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,5120,0.018379555808173288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,5120,0.04423022270202637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,3584,0.021128889587190416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,5120,0.05408622158898247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,4096,0.01594044433699714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,4096,0.03711377912097507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,10240,0.05892355574501885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,3072,0.01927911076280806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,3584,0.014829332629839579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,3584,0.03329244587156508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,2560,0.016671111186345417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,3584,0.04086666636996799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,3072,0.013806222213639153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,3072,0.029647111892700195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,2048,0.014451555079884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,3072,0.03655022382736206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,10240,0.09850577513376872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,2560,0.012631999949614206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,1536,0.011700444751315646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,2560,0.02557777696185642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,2048,0.011375111010339526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,2048,0.02169422143035465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,1024,0.009008888569143083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,2048,0.026991999811596338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,1536,0.010688888529936472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,1536,0.017901332841979135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,768,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,1536,0.023021333747439917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,1024,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,6144,0.06264711088604398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,512,0.006317333214812809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,1024,0.019034667147530448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,768,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,4096,0.04523644513554043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,768,0.01680888897842831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,256,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,512,0.00868800034125646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,512,0.014923555983437432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,128,0.0042239998777707415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,256,0.008967999782827165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,256,0.01312711089849472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,128,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,2560,0.0313937763373057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,64,0.0038533334930737815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1536,32,0.0038888889054457345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1536,128,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,64,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,65536,0.22438044018215605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,32,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,1024,0.013968000809351603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,768,0.009309333231714036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,51200,0.20120533307393393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,65536,0.17188444402482775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,512,0.010256888965765635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,16384,0.06574489010704888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,65536,0.392173343234592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,51200,0.13679556051890054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,51200,0.3098737663692898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,12288,0.05275999837451511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,16384,0.10308977630403306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,65536,0.5556000073750814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1536,128,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,10240,0.041215999258889094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,16384,0.1478053331375122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,51200,0.43132532967461484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,12288,0.03666311171319749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,12288,0.07688533597522311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,8192,0.03261244297027588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,10240,0.03234222200181749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,12288,0.1141093307071262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,10240,0.06514666477839152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,7168,0.02808800008561876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,10240,0.0958168903986613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,8192,0.02422222163942125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,8192,0.052721778551737465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,6144,0.02444177700413598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,8192,0.07894755734337701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,7168,0.02195377813445197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,7168,0.046872887346479625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1536,256,0.008379555410808986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,5120,0.020790222618314955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,7168,0.06981777482562594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,6144,0.04092622134420607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,4096,0.01796444422668881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,5120,0.01763199932045407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,6144,0.06175733274883694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,16384,0.04643733302752177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,5120,0.035192890299691096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,3584,0.016200888488027785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,5120,0.052961776653925575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,4096,0.01550222271018558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,3072,0.01479644411140018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,4096,0.029350221157073975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,3584,0.014554666148291694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,3584,0.026145777768558923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,4096,0.044122666120529175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,2560,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,3584,0.039677331844965615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,3072,0.01348088930050532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,2048,0.010668444136778513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,2560,0.01241422196229299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,2560,0.02054577734735277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,2560,0.030588444736268785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,1536,0.007712889048788283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,2048,0.017363554901546903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,2048,0.011200889117187925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,2048,0.027451554934183758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,6144,0.019926221834288705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,1536,0.010335110955768162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,1024,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,1536,0.014553777045673795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,768,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,1536,0.022669333550665114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,1024,0.009337777892748514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,512,0.00452533322903845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,1024,0.011927111281288994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,1024,0.01864622202184465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,768,0.010055999788973067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,256,0.003572444534964032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,512,0.008678221868144141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,512,0.008617777791288164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,512,0.014797333214018079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,128,0.003208000005947219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,256,0.008282666405042013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,256,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,3072,0.03532088796297709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,64,0.003000000077817175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,256,0.012821333275900947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,128,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,128,0.007341333561473423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,128,0.01035377797153261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,64,0.007050666544172499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,65536,0.20137777593400744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,32,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,51200,0.13679911030663386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,65536,0.17011022567749023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,1024,768,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,65536,0.34447466002570254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,16384,0.0546960002846188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,1024,768,0.016521778371598985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,65536,0.4107360045115153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,51200,0.27433599366082084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,1024,3072,0.023416888382699754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,12288,0.03660444418589274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,16384,0.04605866803063286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,16384,0.09043288893169826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,51200,0.3255244361029731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,10240,0.03156177865134345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,16384,0.11330489317576091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,12288,0.06880178054173787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,8192,0.02383199996418423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,10240,0.031375110149383545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,12288,0.08716177940368652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,10240,0.05749689208136665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,7168,0.022898667388492163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,8192,0.02438933319515652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,8192,0.046987556748920016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,8192,0.06085511048634847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,6144,0.018586667047606576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,7168,0.021565333008766174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,7168,0.041908442974090576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,7168,0.05405866768625048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,5120,0.016404444972674053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,51200,0.13593599531385633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,6144,0.03677599959903293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,6144,0.04805066519313388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,5120,0.03184177809291416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,5120,0.017267554998397827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,1024,32,0.003156444471743372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,4096,0.013808888693650564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,5120,0.040997332996792264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,12288,0.035940445131725736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,3584,0.012156444291273752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,4096,0.01531555586391025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,4096,0.026111111044883728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,3072,0.011000889043013254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,4096,0.03478666808870103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,3584,0.014335999886194864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,10240,0.0733955568737454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,3584,0.02369511127471924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,2560,0.009732444253232744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,3584,0.03170311119821336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,3072,0.013419555293189155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,2048,0.008286221987671321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,3072,0.021055110626750525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,2560,0.01239111108912362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,3072,0.028016000986099243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,2560,0.018908444378111098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,1536,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,2560,0.024671999944580927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,2048,0.01108266661564509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,2048,0.016134222348531086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,6144,0.019500444332758587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,1024,0.005443555613358815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,1536,0.010401777923107147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,2048,0.021746666895018682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,1536,0.013461333182122974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,1024,0.009445333646403419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,1024,0.010688888529936472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,512,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,1024,0.01571466690964169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,768,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,768,0.009458666874302758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,256,0.0035377778112888336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,768,0.01406133340464698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,512,0.008609777523411645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,512,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,128,0.0028533333291610083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,512,0.012757333616415659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,256,0.008289777570300633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,256,0.01147822207874722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,64,0.0028355556229750314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,768,128,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,128,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,128,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,64,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,65536,0.16082666979895696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,768,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,768,1536,0.018567999203999836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,51200,0.11503022246890598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,65536,0.12520710627237955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,65536,0.30143555005391437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,16384,0.04208711120817396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,51200,0.10038577848010594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,51200,0.2376035584343804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,65536,0.40224801169501406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,256,0.007012444237867991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,12288,0.030572444200515747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,16384,0.03515377640724182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,51200,0.32001511255900067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,10240,0.026142223013771906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,12288,0.0273964438173506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,16384,0.11232888698577881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,12288,0.060079110993279346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,8192,0.02029600077205234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,12288,0.08554577827453613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,10240,0.024103111690945093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,768,32,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,10240,0.05068888929155138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,7168,0.017543110582563613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,8192,0.01789777808719211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,10240,0.07228622171613905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,8192,0.041489776637819074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,6144,0.015438222222858004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,8192,0.05969422393374973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,7168,0.016397333807415433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,7168,0.03717777795261807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,5120,0.013802666631009845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,7168,0.05334844523006015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,6144,0.01519733336236742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,6144,0.03272533416748047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,4096,0.011412444214026133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,6144,0.04638311266899109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,5120,0.013463999662134381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,5120,0.027813331948386297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,3584,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,5120,0.040758222341537476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,4096,0.012085333466529846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,16384,0.07970488733715482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,3072,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,4096,0.023257777094841003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,768,32,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,4096,0.034146666526794434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,3584,0.011582222249772815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,2560,0.008655110994974772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,3584,0.021088000800874498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,3072,0.010739555789364709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,3584,0.030459556314680312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,2048,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,2560,0.01034044474363327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,3072,0.027377777629428442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,2560,0.016878222425778706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,1536,0.0063822223908371395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,2048,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,2048,0.014482667048772177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,1024,0.005094222310516569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,2048,0.021195555726687115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,1536,0.00906666699382994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,1536,0.012080888781282636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,768,0.004453333301676644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,1536,0.018238221605618794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,1024,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,1024,0.009649777577983009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,512,0.003567111161020067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,1024,0.015327110886573792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,768,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,768,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,256,0.0031991110493739447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,768,0.013797333670987023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,512,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,512,0.012089778151777057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,128,0.0027297778675953546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,256,0.006654222392373615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,256,0.010737777584128909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,3072,0.019253333409627277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,64,0.0025048889219760895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,128,0.007428444094128079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,128,0.006007110906971826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,512,32,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,128,0.009007111191749573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,64,0.006055999961164262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,512,32,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,65536,0.12411999702453613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,51200,0.08389244476954143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,65536,0.25713690121968585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,65536,0.3896000120374892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,51200,0.09940889146592881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,16384,0.03095910946528117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,51200,0.203348437945048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,256,0.007336888876226213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,12288,0.024559110403060913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,51200,0.30695999993218315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,16384,0.03468622101677789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,16384,0.06753333409627278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,10240,0.02000799940692054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,512,2560,0.02405244443151686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,16384,0.10835911167992486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,12288,0.026703110999531213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,12288,0.05192088749673632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,8192,0.01575555569595761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,65536,0.10674399799770778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,10240,0.023610666394233704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,12288,0.08277422189712524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,7168,0.014491554763582019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,8192,0.017317333155208163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,8192,0.03616355525122748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,10240,0.06958488623301189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,6144,0.0120328888297081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,8192,0.05751199854744805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,7168,0.016131555040677387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,7168,0.03203377789921231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,5120,0.01037688884470198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,6144,0.014913777510325113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,7168,0.05145777927504646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,512,512,0.007603555917739868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,6144,0.04544622368282742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,5120,0.013361777696344586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,5120,0.024284443921513025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,5120,0.03786399960517883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,4096,0.012079111403889127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,4096,0.0203448881705602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,4096,0.03233955634964837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,3584,0.011505777637163797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,3584,0.018623999423450895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,3584,0.029184887806574505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,10240,0.04389333393838671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,3072,0.01681422193845113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,2560,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,3072,0.010849778023031024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,3072,0.02641600039270189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,2048,0.004857777721352047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,2560,0.010399111443095738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,2560,0.015011555618709989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,2560,0.023257777094841003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,4096,0.007691555553012424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,1536,0.004193777839342753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,2048,0.009704888694816166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,2048,0.020503999458418954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,1024,0.003568888952334722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,1536,0.008981333010726506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,1536,0.010939555863539377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,1536,0.01778133379088508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,768,0.0031644445326593188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,1024,0.008345777789751688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,1024,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,3072,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,1024,0.01499288943078783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,512,0.002817777709828483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,768,0.008117333054542542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,768,0.013454222016864352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,512,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,512,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,256,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,512,0.012068444656001197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,6144,0.028439998626708984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,2048,0.012847999731699625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,256,0.007336888876226213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,256,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,256,0.010710222025712332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,3584,0.007178666690985362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,128,0.007320000065697565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,32,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,128,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,256,128,0.008646222452322641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,65536,0.06775289111667208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,64,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,256,32,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,256,768,0.008214222060309516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,51200,0.052573333183924355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,65536,0.12302133772108291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,65536,0.22175999482472739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,16384,0.017870222528775532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,51200,0.09890400038825141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,51200,0.17324177424112955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,65536,0.3858026663462321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,12288,0.013421333498424955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,128,0.0022524444179402455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,16384,0.058343999915652804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,51200,0.30434756808810765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,256,64,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,12288,0.044532444741990834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,12288,0.026038222842746313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,16384,0.1070551143752204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,8192,0.010840000377760993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,12288,0.0815217759874132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,10240,0.023404444257418316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,10240,0.03774666786193848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,7168,0.011445333560307821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,8192,0.01757866640885671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,8192,0.030838221311569214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,10240,0.0687919987572564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,6144,0.010237333675225576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,8192,0.05676800012588501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,7168,0.016221332881185744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,7168,0.027782221635182697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,5120,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,7168,0.0499511096212599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,6144,0.014771555860837301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,6144,0.024367999699380662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,4096,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,5120,0.013282666603724161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,5120,0.02124711043304867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,5120,0.037463999456829496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,16384,0.03418133325046963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,4096,0.01802399920092689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,10240,0.011700444751315646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,4096,0.031911998987197876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,3072,0.006223111102978389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,3584,0.011728000309732227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,3584,0.016568889220555622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,2560,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,3584,0.029112001260121662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,3072,0.014817777607176038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,3072,0.026379555463790894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,2048,0.0047066667013698155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,2560,0.010335110955768162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,2560,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,2560,0.02295644415749444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,1536,0.0041831110914548235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,2048,0.009719111025333405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,2048,0.011735999749766456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,2048,0.020245333512624104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,6144,0.0439280006620619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,1024,0.0035173334181308746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,1536,0.009188444250159794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,1536,0.01000533335738712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,768,0.0031635556370019913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,1536,0.017551110850440133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,1024,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,4096,0.012070222033394707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,512,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,1024,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,1024,0.014782221780882942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,768,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,256,0.0024968888610601425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,768,0.007500444849332173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,512,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,768,0.012768000364303589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,512,0.006559999866618051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,3072,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,512,0.011761777930789523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,256,0.0073928890956772705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,256,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,64,0.0021199999998013177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,32,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,256,0.010039110978444418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2816,128,128,0.007348444726732042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,128,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,65536,0.061596446567111544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2816,128,128,0.00815466625822915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,51200,0.04988977644178602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,32,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,65536,0.2192008892695109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,51200,0.1728817754321628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,12288,0.011021333436171213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,3584,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,16384,0.05817333194944593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,10240,0.010085333552625444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,8192,0.008731555607583787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,7168,0.010610666539933948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,12288,0.04433777928352356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,10240,0.037238223685158625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,6144,0.009679111341635386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,5120,0.008356444537639618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,128,128,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,7168,0.0273982220225864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,4096,0.006979555719428592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,3584,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,6144,0.02439111140039232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,3072,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,5120,0.020971554848882887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,2560,0.005200888961553574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,3584,0.016473778420024447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,128,64,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,2048,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,16384,0.01496088835928175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,1536,0.0038720000949170855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,1024,0.0032008888406885993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,3072,0.014831999937693277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,768,0.002895999906791581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,2560,0.01330844478474723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,512,0.002546666603949335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,256,0.0024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,2048,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,128,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,1536,0.010022222167915767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,1024,0.00831733312871721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,768,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,512,0.00656355544924736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,32,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,256,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,65536,0.061722669336530894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,64,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,32,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,16384,0.012842666771676807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,65536,0.21937510702345106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,12288,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,51200,0.17289422618018258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,4096,0.017924444542990792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,10240,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,8192,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,12288,0.04440977838304308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,7168,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,10240,0.03711466656790839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,6144,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,64,64,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,5120,0.006681777950790193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,7168,0.027440889014138117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,4096,0.008003555238246918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,6144,0.02426133387618595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,128,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,3584,0.0074728892909155945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,64,8192,0.030459556314680312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,5120,0.020992888344658747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,2560,0.005971555494599872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,4096,0.017906667457686532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,3584,0.01650044487582313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,2048,0.00453955555955569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,1536,0.0038577777643998465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,3072,0.014818666709793938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,1024,0.00314666661951277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,2560,0.013443555268976422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,768,0.0028533333291610083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,16384,0.05794399976730347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,2048,0.011712888876597086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,512,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,1536,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,256,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,1024,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,768,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,64,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,8192,0.030532442861133154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,512,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,32,0.0018275556051068837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,256,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,128,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,64,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2816,32,32,0.004963555683692296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,51200,0.04822133315934075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,128,0.0021199999998013177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2816,32,3072,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,16384,1.1092098024156358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,12288,0.8999289406670464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,12288,2.445366117689345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,16384,3.424959182739258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,16384,3.6692657470703125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,12288,2.6008648342556424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,10240,0.6873075697157117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,51200,4.082787407769097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,10240,2.455321841769748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,10240,2.116170671251085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,8192,1.635077370537652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,8192,0.6319964196946886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,7168,0.5024355782402886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,8192,1.8173359764946833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,7168,1.6370497809516058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,6144,0.4281768798828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,6144,1.4233964284261067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,7168,1.4190257390340169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,6144,1.2952889336480033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,5120,0.3618497848510742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,5120,1.371596442328559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,4096,0.951109356350369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,4096,0.8056106567382812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,5120,1.02748531765408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,3584,0.8216684659322103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,3584,0.2644080056084527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,3072,0.625523567199707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,3584,0.8252017762925891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,3072,0.30256978670756024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,3072,0.6061520046657985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,2560,0.20881599850124785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,2560,0.5838969018724229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,4096,0.3025742106967502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,2048,0.45220353868272567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,2048,0.17645511362287733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,51200,10.295147365993923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,1536,0.39496623145209414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,2048,0.42144356833563906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,1536,0.1488062275780572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,1536,0.3335999912685818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,1024,0.2535991138882107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,768,0.20399199591742623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,1024,0.1318488915761312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,512,0.15541777345869276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,768,0.09790844387478298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,256,0.10868889093399048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,512,0.10296533505121867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,768,0.20426577991909453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,128,0.09044177664650811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,512,0.19067466259002686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,256,0.09589244259728326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,1024,0.2738871044582791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,256,0.15437155299716526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,64,0.07617688841289945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,65536,32,0.07431999842325847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,65536,128,0.08342222372690837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,2560,0.5536613464355469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,128,0.14904710981580946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,64,0.15271466308169895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,32,0.14971021811167398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,16384,2.781760957505968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,51200,3.3212142520480685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,65536,51200,15.769605848524305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,65536,4.12675306532118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,16384,0.8319279882642957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,16384,2.912700441148546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,12288,1.9266995324028864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,12288,0.6231128904554579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,10240,0.5855146514044868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,12288,2.1456356048583984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,51200,7.958955552842881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,10240,1.7983288235134547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,10240,1.6948888566758897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,8192,0.4328391022152371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,7168,1.1425537533230252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,8192,1.3383280436197917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,8192,1.671645376417372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,7168,0.38113244374593097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,7168,1.1239395141601562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,65536,9.969228108723959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,6144,0.4131733311547174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,6144,0.9634026421440972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,5120,0.845577769809299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,6144,0.953920046488444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,5120,0.2801084518432617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,4096,0.7542293336656359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,4096,0.23622398906283906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,3584,0.5849235322740343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,5120,0.7994729148017036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,3584,0.21397156185574004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,4096,0.6758515569898816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,3584,0.5516488817003038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,3072,0.18689333067999947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,3072,0.4913484255472819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,2560,0.4237031141916911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,2560,0.16145866447024876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,3072,0.5225804646809896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,51200,11.877716064453125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,2048,0.3225777679019504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,2560,0.4083671040005154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,1024,0.1748595502641466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,2048,0.15805955727895102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,2048,0.353113783730401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,1536,0.11650755670335557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,768,0.13937066660987005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,1536,0.268284453286065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,1024,0.0960728857252333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,768,0.08262311087714301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,1024,0.19934933715396455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,512,0.09564444753858779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,256,0.07104888889524671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,512,0.07387999693552653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,128,0.059395558304256864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,256,0.06411555740568373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,512,0.13946844471825495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,256,0.12213244703080918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,64,0.04933599962128533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,32,0.051651553975211255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,128,0.11596177683936225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,64,0.11740533510843913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,51200,1536,0.26631999015808105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,32,0.11975199646419948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,768,0.16495466232299805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,51200,128,0.06166488594479031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,65536,1.2818942599826388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,51200,65536,15.384455362955729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,51200,0.971135139465332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,16384,0.9619564480251737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,16384,0.2738942305246989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,51200,2.5475554996066623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,12288,0.2119528849919637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,16384,1.0022204717000325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,12288,0.7736604478624133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,10240,0.5206835534837511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,65536,3.2848976982964406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,10240,0.17679556210835776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,12288,0.8235226737128364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,10240,0.5850702391730415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,8192,0.4976186752319336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,8192,0.14818310737609863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,7168,0.3900417751736111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,7168,0.13019643889533147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,51200,3.548952950371636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,6144,0.35608887672424316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,7168,0.3778062131669786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,6144,0.12574488586849636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,6144,0.3234666718377007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,5120,0.28981243239508736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,4096,0.21919911437564424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,4096,0.08346488740709092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,5120,0.27726401223076713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,4096,0.2337884373135037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,3584,0.07447377840677898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,3072,0.15317243999905056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,3584,0.19984711541069877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,2560,0.11960089206695557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,3072,0.06607200039757623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,3072,0.1652079953087701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,65536,5.216580284966363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,2048,0.11099555757310654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,2560,0.057722667853037514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,8192,0.4432986577351888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,2560,0.14060178067949083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,1536,0.08255733384026422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,1024,0.05587377813127306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,2048,0.05957421991560194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,1536,0.04070933328734504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,2048,0.11485155423482259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,1024,0.034487111700905695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,1024,0.07052622238794963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,5120,0.10824088917838202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,768,0.03205422229237027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,3584,0.18254844347635904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,768,0.05633422401216295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,512,0.029422223567962646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,128,0.020953777763578627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,512,0.04811022347874112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,256,0.026587555805842083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,64,0.01755288905567593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,256,0.04354844490687052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,16384,128,0.025693333811230127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,32,0.018422222799725003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,128,0.04116355710559421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,64,0.041448887851503156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,32,0.042119999726613365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,768,0.05055555701255798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,16384,1536,0.0916266640027364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,512,0.03319022059440613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,16384,256,0.026388444834285315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,16384,0.7025840017530652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,65536,1.010160870022244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,51200,0.7649422221713595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,16384,0.6778408686319987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,16384,0.26735999849107533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,12288,0.1628719965616862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,12288,0.5384471151563857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,51200,1.954294204711914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,65536,2.3815360599093967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,10240,0.5018124580383301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,10240,0.15751288996802434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,8192,0.3115786711374919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,12288,0.6578177875942653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,10240,0.40404534339904785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,8192,0.1313813394970364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,7168,0.2775759961869982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,7168,0.1021253334151374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,8192,0.35794933636983234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,6144,0.2537875440385607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,7168,0.30087733268737793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,51200,2.830991956922743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,6144,0.3057937886979845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,5120,0.21104355653127035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,4096,0.16450311077965632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,4096,0.06223289171854655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,3584,0.13719644811418322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,4096,0.16641333368089464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,3072,0.11833422713809544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,65536,4.029788547092013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,3584,0.14377866850958929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,3072,0.05858844518661499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,2560,0.10118133491939968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,3072,0.12557421790228948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,2048,0.08394133382373387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,2560,0.0454657773176829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,2560,0.12006578180525039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,2048,0.03958933220969306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,5120,0.07727466689215766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,2048,0.08854399787055121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,1536,0.032454222440719604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,1536,0.07128266493479411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,768,0.03686844309171041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,1024,0.02639377779430813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,1024,0.053779555691613086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,3584,0.05991466840108236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,512,0.026458667384253606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,768,0.02392533255947961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,768,0.0433573325475057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,256,0.01904444396495819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,512,0.021798221601380244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,128,0.015795555379655626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,512,0.03561066587766012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,256,0.01978133287694719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,64,0.013863999810483722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,256,0.03154311246342129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,6144,0.08948089016808404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,32,0.01404355631934272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,128,0.03048444456524319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,1536,0.06890755229526095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,64,0.03026311265097724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,12288,32,0.030293332205878362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,1024,0.0536391105916765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,65536,0.75510130988227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,51200,0.6199013392130533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,12288,128,0.019682667321628995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,16384,0.1818302207522922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,51200,1.5539991590711806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,16384,0.6575448777940538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,65536,1.98906495836046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,12288,0.3729866610633002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,16384,0.7267777654859754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,12288,0.17926932705773246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,10240,0.3165760040283203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,12288,0.4360035525427924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,10240,0.12045599354637994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,12288,5120,0.2048719988928901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,8192,0.2979493406083849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,10240,0.42079822222391766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,8192,0.09492178095711602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,7168,0.22583466105990938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,7168,0.09947733084360759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,6144,0.2067377832200792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,8192,0.3122524420420329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,6144,0.07618933253818087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,5120,0.17364977465735543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,5120,0.06524266799290974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,7168,0.2400426599714491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,4096,0.13574755191802979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,3584,0.11931288242340088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,5120,0.17602311240302193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,4096,0.054548445675108165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,65536,3.427083545260959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,4096,0.147636440065172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,3072,0.09813511371612549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,3584,0.04785511228773329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,2560,0.07895021968417697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,3072,0.05121333400408427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,2048,0.06692444615893893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,3072,0.10803111394246419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,2560,0.0384479992919498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,51200,2.1622098286946616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,1536,0.05052622159322103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,2560,0.09199466970231797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,2048,0.03376266691419814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,1536,0.027335110637876723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,2048,0.07918933365080091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,768,0.026522666215896606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,1536,0.06024977895948622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,1024,0.023431110713216994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,512,0.02162577708562215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,1024,0.04649600055482653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,768,0.020847110284699336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,256,0.015471110741297403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,768,0.03866488734881083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,512,0.01893422173129188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,128,0.013944889108339945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,6144,0.2078871064715915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,256,0.017284444636768766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,64,0.011349333657158745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,256,0.026825777358478967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,10240,128,0.017208000024159748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,32,0.011618666350841522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,128,0.025983999172846477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,64,0.02604622145493825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,3584,0.12210133340623643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,32,0.02574133376280467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,10240,1024,0.03806844353675842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,65536,0.6471520000033909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,51200,0.5142515500386556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,51200,1.3722399605645075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,65536,1.71015199025472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,16384,0.5552417967054579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,16384,0.15263022316826716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,12288,0.34919201003180605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,12288,0.14093955357869467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,10240,0.3057759867774116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,16384,0.4926755693223741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,51200,1.8642569647894964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,10240,0.11023733350965713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,10240,512,0.03029511041111416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,65536,2.4537927839491105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,8192,0.08034755786259969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,8192,0.22834844059414336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,7168,0.20333688788943818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,6144,0.15651199552747938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,7168,0.075199113951789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,6144,0.06287466817431979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,7168,0.19446667035420737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,5120,0.12979911433325872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,6144,0.17184355523851183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,4096,0.10636533631218804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,5120,0.054089777999454074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,3584,0.09452622466617161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,5120,0.15023733509911433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,3072,0.08460266722573175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,4096,0.1196719937854343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,3584,0.04007110993067423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,3584,0.10117244720458984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,2560,0.0700942211680942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,3072,0.03636799918280707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,12288,0.3395973311530219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,2048,0.05336799886491564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,2560,0.032178666856553816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,3072,0.09274399942821926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,1536,0.04491022229194641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,2560,0.07635377513037787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,10240,0.28887822892930776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,2048,0.02825777729352315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,1536,0.02350933353106181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,2048,0.06247199906243218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,768,0.02310577697224087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,1024,0.01977599991692437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,512,0.017328888177871704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,768,0.017881777551439073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,1024,0.038550221257739596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,256,0.0138586668504609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,512,0.016351110405392118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,4096,0.044481777482562594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,512,0.02535022298494975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,128,0.011020444333553314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,256,0.015211555692884656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,256,0.022364444202846948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,64,0.009303111169073317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,8192,128,0.015212444795502557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,32,0.009958222508430481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,128,0.021672000487645466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,8192,0.23151821560329863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,64,0.02165422174665663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,32,0.021648888786633808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,8192,1024,0.029994666576385498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,1536,0.05014222198062473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,8192,768,0.03148711058828566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,65536,0.6180177794562446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,16384,0.40073333846198184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,51200,0.472267574734158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,16384,0.14732533031039768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,12288,0.2636675569746229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,51200,1.180832862854004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,16384,0.4001457691192627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,10240,0.22627467579311797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,12288,0.10608710845311482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,12288,0.3252631028493245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,10240,0.08956088622411092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,65536,1.899765438503689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,8192,0.18044533994462755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,8192,0.07171555360158284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,10240,0.27250578668382436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,51200,1.6104968388875325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,8192,0.20330399937099883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,7168,0.07092266612582736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,6144,0.16893422603607178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,5120,0.1330204407374064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,6144,0.0572648909356859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,5120,0.04815466536415947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,7168,0.2483582231733534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,4096,0.10583644443088108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,3584,0.0903244415918986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,5120,0.14200622505611843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,4096,0.04001599881384108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,3072,0.07357066869735718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,4096,0.10398311085171169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,3584,0.03716444306903415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,2560,0.06508888800938924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,3584,0.08993511067496406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,3072,0.03215377860599094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,2048,0.04890222350756327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,3072,0.0792142219013638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,2560,0.029431111282772485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,1536,0.03992799917856852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,2560,0.06792000267240736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,2048,0.02532355487346649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,7168,0.16026222705841064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,2048,0.056668447123633496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,1024,0.02681866619322035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,768,0.02180888917711046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,1536,0.02148799929353926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,1024,0.017841777867741056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,512,0.016057777735922072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,768,0.016580444243219163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,65536,2.169668409559462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,256,0.011205332974592844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,768,0.029099557134840224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,512,0.015231110983424716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,128,0.010232888989978367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,256,0.014087999860445658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,256,0.020249777370029025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,64,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,7168,128,0.014094221923086377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,7168,32,0.008984000318580205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,128,0.019699555304315355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,64,0.01959466603067186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,32,0.019560888409614563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,1536,0.04551555713017782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,1024,0.034493333763546415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,6144,0.154968884256151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,65536,0.5184951358371311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,7168,512,0.023177777727444965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,16384,0.2749404377407498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,65536,1.188786718580458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,16384,0.1216684447394477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,51200,1.0725217395358615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,12288,0.2055831087960137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,12288,0.09288889169692993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,16384,0.41391645537482363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,10240,0.2386373413933648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,12288,0.2742346657647027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,8192,0.17244444953070748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,10240,0.07928533024258085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,10240,0.22496443324618867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,65536,1.7791911231146917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,8192,0.062863998942905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,7168,0.14719822671678331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,6144,0.1175902287165324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,7168,0.057690666781531445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,51200,1.508964432610406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,6144,0.05674311187532213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,5120,0.10868177811304729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,7168,0.15279377831353083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,4096,0.0866924458079868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,5120,0.05051733387841118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,3584,0.07448266612158881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,5120,0.11076622539096409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,4096,0.09513422515657212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,3072,0.06303822331958346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,51200,0.4063253402709961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,3584,0.03385511040687561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,2560,0.05374222331576877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,3072,0.029719998439153034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,2048,0.04094311263826158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,3072,0.07433688640594482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,2560,0.06074044439527723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,1536,0.03237866693072849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,2048,0.023870221442646448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,2048,0.04999377661281162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,1024,0.024153777294688757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,1536,0.020245333512624104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,8192,0.2394177648756239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,1536,0.039600888888041176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,768,0.020848888489935134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,1024,0.0173973325226042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,512,0.015475556254386902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,6144,0.13301955329047307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,1024,0.030975110001034204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,768,0.015462223026487561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,256,0.012500444220172035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,4096,0.035678221119774714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,768,0.02558044426971012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,128,0.008990222381220924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,512,0.014136888914638095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,64,0.007667555577225155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,256,0.013213333156373767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,6144,32,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,128,0.01348888874053955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,3584,0.08208711279763116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,128,0.01754044493039449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,64,0.01756088932355245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,6144,2560,0.026559111144807603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,32,0.01750577820671929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,65536,0.4553182390001085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,51200,0.787135124206543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,65536,1.1165635850694444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,512,0.02079288827048408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,6144,256,0.018244443668259513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,51200,0.3514293299780951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,16384,0.27481865882873535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,16384,0.10419022374682957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,12288,0.21749156051211885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,51200,1.1064524120754666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,65536,1.5876106686062281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,16384,0.3162604437934028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,12288,0.0797742207845052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,10240,0.06969244612587823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,8192,0.13649955060746935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,12288,0.3143804338243273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,8192,0.05527822176615397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,10240,0.18421155876583525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,7168,0.12962044609917536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,6144,0.10688177744547527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,8192,0.14987644884321424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,7168,0.0497555567158593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,6144,0.04345688886112637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,4096,0.06201777855555216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,7168,0.14521155092451307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,5120,0.037179556157853864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,6144,0.11241243945227729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,3584,0.058360000451405845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,5120,0.09744889206356472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,4096,0.03148355417781406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,3072,0.048319111267725624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,2560,0.04488089018397861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,3584,0.02919911013709174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,4096,0.0774408910009596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,2048,0.034761776526769005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,3072,0.02567822237809499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,3584,0.06866844495137532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,1536,0.025936888323889837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,2560,0.023297778434223596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,3072,0.0610133343272739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,1024,0.019687111179033916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,2048,0.020209777686331008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,2560,0.053411556614769824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,768,0.015633778439627755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,1536,0.017715555098321702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,2048,0.04383733206325107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,512,0.011356444822417365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,1024,0.014813333749771118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,1536,0.03524355424775018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,256,0.008652444514963362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,768,0.013787555197874704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,1024,0.02674755619631873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,512,0.012725333372751871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,128,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,768,0.022728888524903193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,256,0.011744000017642975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,512,0.01815200017558204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,5120,0.08788000212775336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,256,0.015936000479592215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,32,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,5120,128,0.0120319997270902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,128,0.015455999308162265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,64,0.01550222271018558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,5120,32,0.01540177729394701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,65536,0.3975999885135227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,51200,0.5600986480712891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,10240,0.1846648852030436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,65536,0.7952497800191244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,16384,0.22220177120632598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,51200,0.317300452126397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,16384,0.09158844417995876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,12288,0.1670560042063395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,12288,0.07028088967005412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,5120,64,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,10240,0.1763359970516629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,51200,0.8893653021918403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,12288,0.1996684471766154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,10240,0.061200890276167125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,8192,0.12887110975053576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,8192,0.048892445034450956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,7168,0.0432604451974233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,6144,0.10330044560962254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,8192,0.12295467323727077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,7168,0.10881155729293823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,5120,0.0810391108194987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,6144,0.03881688912709554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,6144,0.09426133500205146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,4096,0.06518577867084079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,5120,0.03208977646297879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,5120,0.07919733391867743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,4096,0.027112000518374976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,4096,0.06459022230572171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,3584,0.024497777223587036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,3072,0.04954666561550564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,3584,0.05711555480957031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,16384,0.2677537865108914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,2560,0.04039644532733493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,3072,0.022328888376553852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,65536,1.251999961005317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,2048,0.03056710958480835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,10240,0.193249781926473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,2560,0.020063999626371596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,1536,0.023194667365815904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,2560,0.04384444488419426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,2048,0.017715555098321702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,1024,0.017714665995703805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,2048,0.03721422288152907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,1536,0.015461333923869662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,768,0.014141332772043018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,1536,0.02966133422321743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,1024,0.013456000222100152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,512,0.01069155583779017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,1024,0.022566222482257422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,768,0.012409778104888068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,3584,0.05872978104485405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,256,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,768,0.019282667173279654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,128,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,256,0.011022222538789114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,512,0.01568888955646091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,256,0.014136888914638095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,64,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,32,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,128,0.010904889139864178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,128,0.013419555293189155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,3072,0.050478223297331065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,64,0.013444444371594323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,4096,7168,0.12265155050489639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,4096,32,0.013223999904261695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,51200,0.26953066719902885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,65536,0.6735431353251139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,51200,0.5121279822455512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,16384,0.20718577173021105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,4096,512,0.011699555648697747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,16384,0.08542578087912665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,12288,0.15981688764360216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,51200,0.824861314561632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,65536,1.0248195860120985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,12288,0.06709244516160753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,10240,0.12750755416022405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,8192,0.10676711135440403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,12288,0.16432266765170628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,10240,0.05699111355675591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,7168,0.09427644146813287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,8192,0.04643822378582425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,10240,0.14238044950697157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,6144,0.08163466718461779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,8192,0.14518045054541692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,7168,0.039749334255854286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,5120,0.06369511286417644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,6144,0.03542044427659776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,4096,0.04181333382924398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,5120,0.03083199924892849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,65536,0.3441920015547011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,6144,0.08549866411421035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,3584,0.04141511188613044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,4096,0.026022222306993272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,5120,0.09438133239746094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,3584,0.02387911081314087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,2560,0.03107644451989068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,3072,0.02281066609753503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,3584,0.052016887399885386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,2048,0.02539644473128849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,3072,0.04640266630384657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,16384,0.23219556278652617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,2560,0.019928889142142404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,1536,0.01864355636967553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,2560,0.04028888874583774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,2048,0.017082666357358296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,1024,0.01456622282663981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,1536,0.014531556102964612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,2048,0.033815999825795494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,768,0.012044444680213928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,1024,0.012990222209029727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,1536,0.027445332871543035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,512,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,1024,0.021037333541446265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,768,0.012328888807031842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,256,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,768,0.017840888765123155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,512,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,7168,0.10056355264451768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,128,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,256,0.01068800025516086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,256,0.012828444441159567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3584,128,0.010622222390439777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,32,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,128,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,3072,0.03459199931886461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,4096,0.05848711066775852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,64,0.012383999923865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,32,0.01239999963177575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,65536,0.2871902253892687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,65536,0.5383982128567165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,51200,0.5306888686286079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,51200,0.23241244422064888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,16384,0.1604293319914076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,12288,0.12235111660427517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,16384,0.09174488650427924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,51200,0.7052488856845431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3584,64,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,65536,0.9134053124321831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,12288,0.1518204477098253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,10240,0.10842933257420857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,8192,0.07871644364462958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,10240,0.05002133382691277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,8192,0.038520889149771795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,7168,0.07597511344485812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,10240,0.12064533763461643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,8192,0.10220622354083592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,6144,0.06154844495985243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,7168,0.03436355458365546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,5120,0.0539173318280114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,6144,0.030235555436876085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,7168,0.08851822217305501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,6144,0.07647199763192071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,4096,0.03992799917856852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,5120,0.02679644525051117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,3584,0.0377511117193434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,5120,0.06487644380993314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,4096,0.023052444060643513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,3072,0.030889779329299927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,4096,0.05297599898444282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,16384,0.1968604458702935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3584,512,0.014459555347760519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,2560,0.0270462218258116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,3584,0.04717422193951077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,12288,0.06512444549136691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,3072,0.01901422275437249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,2048,0.022085333863894146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,1536,0.018170666363504197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,3072,0.04239822096294827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,2560,0.017222222354676988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,1024,0.01348533315791024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,2048,0.01554844445652432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,2560,0.036553776926464505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,1536,0.013935110635227628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,2048,0.030749334229363337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,1024,0.012091555529170565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,1536,0.024503111839294434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,512,0.008914666871229807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,1024,0.018982221682866413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,768,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,256,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,512,0.010583110981517367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,768,0.016199999385409884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,128,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,512,0.013391999734772576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,64,0.004994666824738185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,128,0.010051555931568146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,256,0.011759111450778114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,128,0.011398221883508893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,64,0.011354666617181567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,3584,0.02089866664674547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,3072,32,0.011343110766675739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,65536,0.5849964353773329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,65536,0.27071555455525714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,768,0.011591110792424945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,51200,0.4009111192491319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,16384,0.14212800396813288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,16384,0.06839111116197374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,65536,0.7253537707858615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,51200,0.23235556814405653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,51200,0.5825075573391384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,3072,256,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,10240,0.09179733196894328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,12288,0.053435554107030235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,3072,32,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,8192,0.07557777563730876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,10240,0.04623644550641378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,12288,0.12804178396860758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,7168,0.06699466705322266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,10240,0.10879111289978027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,8192,0.0362737774848938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,6144,0.054791112740834556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,7168,0.03300977746645609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,5120,0.04458399944835239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,7168,0.07738488912582397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,4096,0.03734222385618422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,5120,0.02866222129927741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,6144,0.06775555345747206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,3584,0.033124443557527326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,5120,0.05664088990953234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,4096,0.02293333411216736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,3072,0.027161777019500732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,4096,0.046590222252739795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,3584,0.041962666643990405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,12288,0.1149137814839681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,2560,0.023050665855407715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,3072,0.018368888232443068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,2048,0.018616888258192275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,16384,0.16862222883436415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,3072,0.03728711273935106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,1536,0.01517599986659156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,2560,0.01686044368478987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,2048,0.014908444550302295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,2560,0.0321084459622701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,1024,0.011066666907734342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,1536,0.012431999875439538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,2048,0.02702666653527154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,768,0.009000000026490953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,1536,0.022152000003390845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,1024,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,512,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,6144,0.029319110843870375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,768,0.010613333019945355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,256,0.005423111220200856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,768,0.014508444401952954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,512,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,512,0.011907555162906647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,128,0.005200888961553574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,256,0.009436444275909001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,64,0.004514666481150521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,256,0.0107342220014996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,3584,0.02016622159216139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2560,32,0.004607999904288186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2560,128,0.009646221995353699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,128,0.010307555397351583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,64,0.010110222631030613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,32,0.010600888894663917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,65536,0.43880889150831437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,65536,0.23983200391133627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,51200,0.28817688094245064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,16384,0.10756444268756443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,8192,0.08719555536905925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2560,1024,0.016914667354689706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,51200,0.1999057796266344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,12288,0.08366844389173721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,16384,0.057745774586995445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,51200,0.46075020896063906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,16384,0.14258400599161783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,12288,0.05089066757096184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,8192,0.05089955528577169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,12288,0.10756888654496934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,10240,0.04007200068897671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,10240,0.09147733449935913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,7168,0.047637333472569786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,8192,0.0310533344745636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,6144,0.043403555949529014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,8192,0.07662222120496961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,7168,0.02827022141880459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,5120,0.034240000777774386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,7168,0.06571111414167616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,6144,0.02516799999607934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,4096,0.028373334142896865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,6144,0.056141330136193164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,5120,0.02206666602028741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,3584,0.023923556009928387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,3072,0.02158222264713711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,4096,0.01902755598227183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,4096,0.039149334033330284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,3584,0.01742577718363868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,2560,0.018066666192478605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,3072,0.01626488897535536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,3584,0.03535555468665229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,2048,0.01639111174477471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,65536,0.583113776312934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,2560,0.01482311056719886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,1536,0.013058666553762225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,10240,0.06755644745296903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,2048,0.013545778062608508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,1024,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,1536,0.01205599970287747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,2048,0.023016888234350417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,768,0.007703999678293864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,1536,0.018912000788582694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,1024,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,512,0.006178666734033161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,1024,0.014811555544535318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,768,0.010040889183680216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,256,0.005134222408135732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,512,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,128,0.004219555606444676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,512,0.01073155552148819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,256,0.009371555513805812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,64,0.0038382220599386427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,128,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,128,0.009297777381208207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,2048,32,0.0038906666967603895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,5120,0.048030220799975924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,64,0.009032888544930352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,32,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,3072,0.03156444430351257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,65536,0.31872712241278756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,2560,0.027067555321587458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,65536,0.1738497813542684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,51200,0.2265982098049588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,16384,0.0735351112153795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,51200,0.1410693327585856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,65536,0.4777741962009006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,12288,0.05465777715047201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,10240,0.046563555796941124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,16384,0.04679822259479099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,2048,768,0.012864888542228274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,12288,0.03580888774659898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,51200,0.3704506556193034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,16384,0.12066133817036946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,8192,0.041313777367273964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,10240,0.03246488836076524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,12288,0.09490844276216294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,7168,0.03594933284653558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,8192,0.025008888708220586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,6144,0.029021332661310833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,10240,0.07541333304511176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,5120,0.02683288852373759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,8192,0.06208533048629761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,7168,0.022871999277008906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,4096,0.021446221404605444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,6144,0.020297777321603563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,7168,0.054416000843048096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,5120,0.018216000662909616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,3584,0.019425777925385367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,6144,0.047784000635147095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,4096,0.015673778123325773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,5120,0.04036800066630045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,3584,0.014693333042992486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,4096,0.034120000070995755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,2560,0.014541332920392355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,3584,0.030625777112113103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,2048,0.012418666647540199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,2560,0.012439111040698158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,2560,0.023221333821614582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,1536,0.010025777750545079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,2048,0.011413333316644033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,2048,0.020107555720541213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,1024,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,1536,0.01055555542310079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,2048,256,0.009004444711738164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,1536,0.016700444949997794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,1024,0.009642666412724389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,1024,0.012960000170601739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,512,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,768,0.009291555318567488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,768,0.011170667078759937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,256,0.004263111286693149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,512,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,3072,0.017086222767829895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,512,0.009509333305888707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,128,0.003850666599141227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,256,0.008272888759771982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,3072,0.013624889155228933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,256,0.008536888493431939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1536,128,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,32,0.0035439998739295532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,3072,0.026895999908447266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,128,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,64,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,65536,0.2273182339138455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,65536,0.16555821895599365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,51200,0.1900391048855252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,768,0.0068400001360310455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,51200,0.1312613354788886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,65536,0.37404266993204754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,12288,0.04363822274737888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,16384,0.04467288984192742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,51200,0.2923804389105903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,10240,0.036638223462634616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,12288,0.03570400012864007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,8192,0.028040889236662123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,12288,0.07306666506661309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,10240,0.03085422184732225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,10240,0.061477336618635386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,8192,0.023541332946883306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,8192,0.05115555392371284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,6144,0.021670222282409668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,7168,0.02130577796035343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,7168,0.04459644357363383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1536,32,0.008036444584528605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,6144,0.019211555520693462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,6144,0.039431111680136785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,4096,0.015135111080275642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,16384,0.05876355701022678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,5120,0.03345155715942383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,4096,0.015173332558737861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,4096,0.027767111857732136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,16384,0.09983911116917928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,3072,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,3584,0.014144000079896716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,3584,0.025303999582926433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,2560,0.01015466699997584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,7168,0.024597333537207708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1536,64,0.0034977777136696708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,3072,0.022491556074884202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,2048,0.008779555559158325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,2560,0.012141333685980903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,1536,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,2560,0.019885333047972787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,2048,0.011051555474599203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,1024,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,2048,0.016917333006858826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,1536,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,768,0.0051857779423395796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,1536,0.014302222265137566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,1024,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,512,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,768,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,768,0.009949333137936061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,3584,0.013280889226330651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,256,0.0034826666944556763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,512,0.008398222426573435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,512,0.008355555435021719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,128,0.0031973332580592898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,256,0.008047111332416534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,256,0.0074106670088238185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,64,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,128,0.008017777568764156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,3072,0.013108444710572561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,32,0.0031475555151700974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,128,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,65536,0.17917777432335746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,64,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,32,0.00700800038046307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,1024,5120,0.017433777451515198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,51200,0.16118133068084717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,16384,0.04956444435649448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,65536,0.16289244757758245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,1024,5120,0.017172444197866652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,65536,0.32734754350450307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,1024,1024,0.011408000356621213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,12288,0.03843555516666836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,51200,0.12971733676062688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,10240,0.03101955519782172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,16384,0.04403644469049242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,51200,0.25741510921054417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,8192,0.023943111300468445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,16384,0.08621866835488214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,10240,0.03042666779624091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,7168,0.021184888150956895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,10240,0.05465333329306709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,6144,0.018516444497638278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,7168,0.021433777279324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,8192,0.044621331824196704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,5120,0.016214221715927124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,7168,0.040241777896881104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,6144,0.019088000059127808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,4096,0.013447110851605734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,6144,0.03496177660094367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,5120,0.01698755555682712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,3584,0.011026666396194033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,5120,0.029931555191675823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,4096,0.01517955627706316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,3072,0.00960177762640847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,4096,0.025261332591374714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,3584,0.014133334159851074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,2560,0.00813066628244188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,3584,0.02265600032276577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,2048,0.007326222128338284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,3072,0.02036266691154904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,2560,0.01204355557759603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,1536,0.006221333311663733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,2560,0.017906667457686532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,12288,0.03423644436730279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,2048,0.011032889286677042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,1024,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,12288,0.06460977925194634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,768,0.004323555363549126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,1536,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,8192,0.023524444964196947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,512,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,1536,0.012796444197495779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,1024,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,256,0.003152888889114062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,768,0.009027555584907532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,1024,0.010556444525718689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,512,0.008282666405042013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,256,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,512,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,64,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,256,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,32,0.0028719999310043124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,128,0.008029333419269985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,128,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,64,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,65536,0.16487377219729954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,32,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,768,3072,0.01309511148267322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,51200,0.13221066527896458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,16384,0.03596977723969354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,65536,0.11930844518873428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,65536,0.28490132755703396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,51200,0.09579022063149346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,2048,0.015448888142903646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,12288,0.026475555366939966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,16384,0.032891554964913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,51200,0.2246088981628418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,12288,0.025810667210155066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,16384,0.07714311281840007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,8192,0.01590044465329912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,12288,0.056644446320003934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,10240,0.023200000325838726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,768,128,0.003173333282272021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,7168,0.0140684445699056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,10240,0.048354665438334145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,6144,0.012366222010718452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,8192,0.01720177796151903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,5120,0.011014222270912595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,7168,0.01611288885275523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,6144,0.014518222875065274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,7168,0.03553422292073568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,6144,0.031312889522976346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,5120,0.013293333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,4096,0.009292444421185387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,3584,0.00831200016869439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,5120,0.026364444030655756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,4096,0.011777777638700275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,3584,0.01112711098459032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,4096,0.022266666094462078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,3584,0.02032888929049174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,2560,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,3072,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,10240,0.021077333225144282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,3072,0.018195556269751657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,2048,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,2560,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,2560,0.01607733302646213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,2048,0.009665778113736046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,768,768,0.009155555731720394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,2048,0.013909333282046847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,1024,0.00416355538699362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,1536,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,768,0.003664888855483797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,1536,0.011695110963450538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,1024,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,512,0.0031831111345026228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,1024,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,768,0.008023999631404877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,256,0.0028106667515304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,768,0.008572444319725037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,512,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,128,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,512,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,3072,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,256,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,64,0.002487999904486868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,256,0.006314666734801398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,512,128,0.007425777614116669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,128,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,65536,0.08511200216081406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,64,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,32,0.006131555471155379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,51200,0.06332711378733318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,1536,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,16384,0.024403555525673762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,65536,0.11797955301072861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,51200,0.09532177448272705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,12288,0.01812977757718828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,51200,0.19144622484842935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,16384,0.06387199958165486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,10240,0.018760000665982563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,12288,0.025205334027608235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,12288,0.04885422190030416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,10240,0.022451554735501606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,512,32,0.00250133333934678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,7168,0.0140346669488483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,8192,0.01721244388156467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,8192,0.034062223301993474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,6144,0.01204533295498954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,7168,0.015784000356992085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,7168,0.030656887425316706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,6144,0.014494222071435718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,6144,0.026770666241645813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,65536,0.24267021814982095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,5120,0.01313777764638265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,4096,0.010108444425794814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,16384,0.032427555984920926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,3584,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,5120,0.023316444622145757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,4096,0.011719110939237805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,3072,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,3584,0.0110871113008923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,4096,0.01954844428433312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,8192,0.015431111057599386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,2560,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,3584,0.017904000149832833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,2048,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,3072,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,2560,0.010054222411579555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,3072,0.01608533329433865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,10240,0.041420443190468684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,1536,0.0042053332759274375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,2048,0.009602666729026372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,1024,0.0034764446318149567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,2560,0.014378666877746582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,768,0.003206222214632564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,2048,0.012485333614879184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,1024,0.008307555483447181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,1536,0.010574222438865237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,512,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,768,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,256,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,5120,0.01056977775361803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,768,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,512,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,128,0.0022373333987262514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,512,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,256,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,64,0.0021395554973019496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,512,8192,0.0401760008600023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,128,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,256,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,128,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,65536,0.06515733400980632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,64,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,32,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,51200,0.05188444587919447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,65536,0.11744177341461182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,16384,0.014633778068754407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,65536,0.20578400293986002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,51200,0.09510844283633763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,12288,0.012116444607575735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,256,1536,0.00903555585278405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,51200,0.16078666845957437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,16384,0.032227555910746254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,10240,0.011076444553004371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,16384,0.05418133404519823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,256,1024,0.008632000121805403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,12288,0.02475111186504364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,8192,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,10240,0.023189332750108507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,7168,0.011215111447705163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,10240,0.035050666994518705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,8192,0.017248888810475666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,6144,0.011477332976129321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,7168,0.015835555063353646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,8192,0.02906399965286255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,256,32,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,7168,0.026013332936498854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,6144,0.014507555299335055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,4096,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,6144,0.022967111733224656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,5120,0.01312711089849472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,3584,0.006999111009968652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,4096,0.012088889049159156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,5120,0.020067556036843192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,3072,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,3584,0.011385777758227454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,2560,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,3584,0.015632000234391954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,3072,0.010846222440401712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,2048,0.004577777865860197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,3072,0.014210666219393412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,2560,0.010060444474220276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,1536,0.004019555532270008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,2560,0.012799111505349478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,2048,0.009654222263230218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,1024,0.0035004446076022256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,2048,0.011191111471917896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,1536,0.00903911143541336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,12288,0.04153688748677572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,768,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,1536,0.009659555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,512,0.002834666727317704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,1024,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,1024,0.007962666451931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,768,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,5120,0.0090506664580769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,512,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,128,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,512,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,256,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,64,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,32,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2561,128,128,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,128,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,4096,0.017108443710539076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,65536,0.058655997117360435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,64,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,51200,0.04607199960284763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,32,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,16384,0.012753778033786349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,12288,0.011591110792424945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,10240,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,65536,0.20459111531575522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,51200,0.16067822774251303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,8192,0.00906755526860555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,12288,0.04156000084347195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,7168,0.009823111196359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,10240,0.03473155697186788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,6144,0.008456000023418004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,5120,0.0083155557513237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,8192,0.028759999407662287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,128,256,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,768,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,6144,0.022916444473796423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,3584,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,3072,0.005736888696750005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,5120,0.019951999187469482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,2560,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,128,256,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,4096,0.016972444123691983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,2048,0.004558222161398994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,1536,0.0038844446341196695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,3584,0.015605333778593274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,1024,0.0032017777363459268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,3072,0.014095111025704278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,768,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,2560,0.0127697777416971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,512,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,2048,0.011074666347768573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,1536,0.009681777821646797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,256,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,1024,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,128,0.0021235555824306277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,768,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,16384,0.05413777629534403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,64,0.001976888833774461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,32,0.00208177769349681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,256,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,128,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,65536,0.0565884444448683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,64,0.005097777893145879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,51200,0.044312887721591525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,32,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,16384,0.0122079998254776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,12288,0.010139555566840703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,65536,0.2047226693895128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,16384,0.054153776831097074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,51200,0.1603191163804796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,10240,0.009130666653315226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,12288,0.04126311010784573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,8192,0.008362666600280339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,7168,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,6144,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,10240,0.03491733471552531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,8192,0.028664890262815688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,5120,0.010344888601038191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,4096,0.008624888956546783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,6144,0.0229777776532703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,3584,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,512,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,5120,0.0199608885579639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,4096,0.016864000095261466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,3584,0.015785778562227886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,64,7168,0.025787555509143408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,3072,0.014159111513031853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,2048,0.004564444637960858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,1536,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,2560,0.01276266657643848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,1024,0.0031991110493739447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,2048,0.011158222125636207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,768,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,1024,0.008096888661384583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,512,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,256,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,7168,0.025756445195939805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,128,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,512,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,64,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,256,0.005636444522274866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,32,0.001867555495765474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,128,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,2560,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,64,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,64,4096,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,32,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,1536,0.009704888694816166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2561,32,768,0.007272889216740926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2561,32,3072,0.0070168889231152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,16384,1.031874656677246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,16384,3.1598932478162975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,16384,3.758098602294922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,51200,3.728401607937283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,12288,2.544125239054362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,16384,3.0577030181884766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,12288,2.5088204277886286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,12288,0.7191644244723849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,10240,0.609515560997857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,12288,2.3155528174506292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,10240,2.0971785651312933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,10240,2.084409713745117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,10240,1.9237289428710938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,8192,1.7683021757337782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,8192,0.5224506590101454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,8192,1.6367350684271917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,51200,10.61198255750868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,8192,1.5358702341715496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,7168,0.4582595295376248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,7168,1.3912622663709853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,6144,0.3953884442647298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,7168,1.3683119879828558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,7168,1.3178284962972004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,6144,1.214355574713813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,6144,1.209546618991428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,5120,0.3425911002688938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,5120,1.2000853220621746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,6144,1.1371013853285048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,4096,0.8859830962287055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,5120,0.9845075607299805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,4096,0.28418132993910045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,5120,0.9555848439534506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,4096,0.8027822176615397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,3584,0.6447475751241049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,3584,0.2530844476487902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,3584,0.6864426930745443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,4096,0.728824880388048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,3072,0.5118035740322536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,3584,0.6562879880269369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,3072,0.2270266744825575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,3072,0.6211875809563531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,2560,0.5318737559848362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,51200,14.794718424479166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,2560,0.19606310791439482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,3072,0.5601280000474718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,2560,0.5286702050103081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,2048,0.45228264066908097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,2560,0.47327200571695965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,2048,0.16677688227759468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,1536,0.26563821898566353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,2048,0.4429608980814616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,51200,10.304935031467014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,2048,0.390938679377238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,1536,0.13467378086513943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,1536,0.4499084684583876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,768,0.1624506711959839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,1536,0.30955200725131565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,1024,0.11060355769263373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,1024,0.2503155602349175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,512,0.12698754999372694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,768,0.09499466419219971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,1024,0.2614702118767632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,768,0.20298666424221465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,256,0.08124978012508817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,512,0.08355644676420425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,512,0.1696773370107015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,768,0.1991200049718221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,128,0.07099199957317777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,512,0.16204710801442465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,64,0.06422399812274508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,256,0.12054755952623154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,128,0.07158044311735365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,256,0.15431377622816297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,128,0.14353244834476048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,32,0.06303111049864027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,65536,128,0.09396711322996353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,65536,1024,0.1870488855573866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,64,0.14647644095950657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,65536,32,0.1464302274915907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,65536,256,0.07641600237952338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,51200,3.020243538750542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,65536,3.7334247165256076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,51200,7.952495998806423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,16384,3.196255154079861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,16384,3.0041359795464406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,16384,0.7868755658467611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,65536,10.220544603135851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,12288,2.154703140258789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,12288,0.5907635688781738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,12288,2.108152813381619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,51200,11.806680467393663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,16384,2.410405264960395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,51200,8.184249877929688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,10240,0.48729244867960614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,12288,1.6039369371202257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,10240,1.6887440151638453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,10240,1.6541306177775066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,8192,1.2799751493665907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,8192,0.40322399139404297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,8192,1.3405031628078883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,10240,1.4709475835164387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,7168,1.1281359990437825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,65536,14.831895616319445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,7168,0.35733689202202695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,8192,1.134950213962131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,7168,1.0926088756985135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,65536,10.306663513183594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,6144,1.1111022101508248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,7168,1.0330897437201607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,5120,0.8255759874979655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,6144,0.9730728997124566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,5120,0.30363467004564076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,5120,0.7733724382188586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,6144,0.8611040115356445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,4096,0.25403377744886607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,4096,0.7456062104966906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,6144,0.30844089719984263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,5120,0.6828959782918295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,4096,0.6620737711588541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,3584,0.227363559934828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,4096,0.5850844383239746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,3584,0.5702533192104763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,3072,0.4829004605611165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,3072,0.18196354971991646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,3584,0.500048902299669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,2560,0.43302755885654026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,3072,0.4848151206970215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,3072,0.4362960126664903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,2560,0.157368885146247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,2048,0.34807022412618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,2560,0.4048115677303738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,2048,0.13120888339148626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,2560,0.3708382182651096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,3584,0.4944586753845215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,1536,0.22509688801235625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,2048,0.3722879886627197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,1024,0.16689244906107584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,1536,0.10609600279066299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,1536,0.28843288951449925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,768,0.13222132788764104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,1024,0.08500444226794773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,1536,0.24723821216159395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,1024,0.1854462226231893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,512,0.11270666122436523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,768,0.15688888231913248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,1024,0.19631644090016684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,256,0.07974666357040405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,768,0.15381866031222874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,512,0.06787555747561984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,512,0.1355795595380995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,128,0.05530488822195265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,512,0.12313066588507758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,256,0.060007108582390674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,256,0.11940622329711914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,256,0.09741066561804877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,64,0.04584266742070516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,51200,32,0.059608002503712974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,128,0.05791733662287394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,128,0.1134986612531874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,2048,0.30092178450690377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,64,0.11522489123874241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,51200,32,0.11573244465721978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,51200,768,0.07554044326146443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,51200,128,0.0736480024125841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,65536,1.188937822977702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,51200,0.9141866895887586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,51200,2.580211639404297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,16384,1.0361173417833116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,65536,3.2576541900634766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,16384,0.27265776528252494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,12288,0.6276666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,51200,3.732227749294705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,16384,1.113658693101671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,16384,0.7419617970784506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,12288,0.6643715434604222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,12288,0.20108889208899605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,10240,0.4745333459642198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,65536,4.903763665093316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,10240,0.17326933807796904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,10240,0.5342737833658854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,12288,0.5595013300577799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,10240,0.4614906840854221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,65536,3.371018727620443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,8192,0.42109423213534886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,8192,0.1399431096182929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,51200,2.7477974361843533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,7168,0.31768621338738334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,8192,0.4423324531979031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,7168,0.13118844562106663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,8192,0.37777333789401585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,6144,0.2753360006544325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,7168,0.38692887624104816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,6144,0.1094577776061164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,5120,0.2613706588745117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,7168,0.3342711130777995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,6144,0.35076265864902073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,5120,0.09292978048324585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,4096,0.19647022088368735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,5120,0.2670559883117676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,6144,0.3161902162763807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,3584,0.186246222919888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,4096,0.07902933491600884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,5120,0.2459742228190104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,4096,0.2393804391225179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,3584,0.06947466399934557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,4096,0.1968853341208564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,3072,0.146978669696384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,3584,0.19245955679151747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,2560,0.11050133572684394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,3072,0.061936000982920326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,3584,0.17508089542388916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,2048,0.10326577557457818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,3072,0.1609919998380873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,2560,0.055359999338785805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,1536,0.07700888978110419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,2560,0.13740711741977268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,2560,0.13936000400119358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,2048,0.04722222354676989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,2048,0.11317955123053657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,1024,0.05451022254096138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,1536,0.039890666802724205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,2048,0.11071466737323338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,1536,0.09029155307345921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,768,0.04365155431959364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,1024,0.03386755453215705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,1536,0.08965688943862915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,1024,0.06883822547064887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,512,0.03301599952909682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,768,0.03073599934577942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,1024,0.07088444630304973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,768,0.055212444729275174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,256,0.02232355541653103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,512,0.02812533246146308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,768,0.059226665231916636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,128,0.01994222237004174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,512,0.04749866657786899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,256,0.025157334076033697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,512,0.04943644338183933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,64,0.01683911184469859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,256,0.04254577888382805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,16384,32,0.01752444439464145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,16384,128,0.02443644404411316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,256,0.03904177745183309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,128,0.04047022263209025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,128,0.0298942228158315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,64,0.04103555613093906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,16384,3072,0.15757155418395996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,16384,32,0.04107822312249078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,65536,0.8821893268161349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,51200,1.9709981282552083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,16384,0.5846080250210232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,51200,0.6884924570719401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,65536,2.659815894232856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,16384,0.2302266756693522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,51200,2.707405302259657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,12288,0.5079030990600586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,16384,0.678109327952067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,16384,0.5530853271484375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,51200,2.0130257076687283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,10240,0.38698310322231716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,12288,0.5125022464328343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,12288,0.1559786664115058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,65536,3.694830152723524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,12288,0.4196382098727756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,10240,0.41996532016330296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,10240,0.129622220993042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,8192,0.3349653349982367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,10240,0.3719200028313531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,8192,0.10680711269378662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,7168,0.2356648974948459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,8192,0.3254586590660943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,8192,0.2899048858218723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,7168,0.09305066532558864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,6144,0.22431733873155382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,7168,0.31440801090664333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,6144,0.08400355445014106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,7168,0.2785049014621311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,5120,0.20145511627197266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,6144,0.2470568815867106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,6144,0.21983644697401258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,4096,0.15464088651869032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,5120,0.0724195573065016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,5120,0.20022222730848524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,3584,0.12908000416225857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,5120,0.1832782162560357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,4096,0.06128622425927056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,4096,0.17810222837660047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,3072,0.11444889174567328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,4096,0.15279022852579752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,3584,0.14259200625949436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,2560,0.0841288897726271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,3584,0.13429066869947645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,3072,0.048130667871899076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,3072,0.1224222183227539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,2048,0.0773493316438463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,3072,0.12266488869984944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,2560,0.04303288790914747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,2560,0.10609777768452962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,1536,0.058711998992496066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,2048,0.037504888243145414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,2048,0.08720799949434067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,2560,0.10549155871073405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,1024,0.03786311215824551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,2048,0.08752889103359646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,1536,0.0694355567296346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,768,0.03107644451989068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,1536,0.07144888904359606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,1024,0.02462933295302921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,1024,0.053391112221611865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,512,0.02500711050298479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,1024,0.05530933207935757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,768,0.023334221707450017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,3584,0.055249777105119496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,256,0.024832000335057575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,768,0.04704711172315809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,512,0.02103288968404134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,512,0.03478399912516276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,128,0.02100533412562476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,512,0.039846221605936684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,256,0.018888889087571036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,256,0.031173331869973078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,64,0.013543999857372709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,256,0.03205777870284186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,128,0.01906044450071123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,128,0.029904001288943823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,12288,32,0.013487111363146039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,128,0.02443466583887736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,12288,1536,0.030896888838873968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,64,0.029792000850041706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,32,0.029607110553317603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,12288,768,0.04233333468437195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,12288,65536,2.609926223754883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,65536,0.7395404179890951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,51200,0.5937937630547417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,65536,2.0447396172417536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,16384,0.7067591349283854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,51200,1.9866506788465712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,16384,0.1978915532430013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,51200,2.225516425238715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,16384,0.6485448943244086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,12288,0.49718311097886825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,65536,2.829851574367947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,16384,0.4764186541239421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,12288,0.14622399542066786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,12288,0.4067964553833008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,10240,0.33941155009799534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,51200,1.7459422217475042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,12288,0.37581600083245176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,10240,0.11012444231245254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,65536,2.2339155409071183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,10240,0.34388976626926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,8192,0.2864302264319526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,8192,0.08958666854434544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,7168,0.25242845217386883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,10240,0.30823911560906303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,8192,0.32428354687160915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,8192,0.24675112300448945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,7168,0.23428710301717123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,7168,0.25501243273417157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,6144,0.20585955513848198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,5120,0.1396204498079088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,6144,0.18821422259012857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,5120,0.06128355529573229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,5120,0.17042844825320771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,4096,0.13522044817606607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,5120,0.16024000114864773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,4096,0.05211288730303446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,3584,0.11909600098927815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,4096,0.136080887582567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,4096,0.13064711623721653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,3584,0.046289778418011136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,7168,0.07992266946368747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,3072,0.10019110971026951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,3584,0.11976533465915257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,3584,0.11972978379991318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,2560,0.0798471106423272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,3072,0.04116266634729173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,6144,0.07130844725502862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,3072,0.10515111022525364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,2560,0.036826666858461164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,2048,0.07127466466691759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,2560,0.09045955869886611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,3072,0.10564621951844956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,1536,0.048947556151284106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,2560,0.09164533350202773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,2048,0.03175110949410333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,1024,0.0377786656220754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,2048,0.07466044690873888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,1536,0.025416889124446448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,768,0.025832000705930922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,1536,0.058993776639302574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,1024,0.02145599987771776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,1536,0.0636062224706014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,512,0.023003555006451074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,1024,0.0454053348965115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,1024,0.05003822180959913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,768,0.01957066688272688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,768,0.036807109912236534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,256,0.018420444594489205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,768,0.042007999287711255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,512,0.018111111389266122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,512,0.029652443197038438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,128,0.015498666299713982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,6144,0.2089653280046251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,512,0.03517155514823066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,256,0.01646844380431705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,64,0.014116444521480136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,256,0.026418666044871014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,10240,32,0.01461866663561927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,256,0.027699554959932964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,10240,128,0.016228444046444364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,128,0.025454221500290766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,128,0.021314667330847845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,64,0.025457777910762366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,10240,32,0.025196444657113817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,10240,2048,0.0780071086353726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,65536,0.613268428378635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,51200,0.48729419708251953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,16384,0.43090044127570254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,65536,1.6763564215766058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,51200,1.5288711123996312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,16384,0.1589662233988444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,16384,0.49639023674858945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,12288,0.3881857660081651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,51200,1.8039866553412542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,16384,0.43114577399359805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,65536,2.2575181325276694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,12288,0.14095200432671443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,10240,0.2609688970777723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,12288,0.3339946534898546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,12288,0.29993777804904515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,8192,0.20986666944291857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,10240,0.11335644457075332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,65536,1.872969733344184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,10240,0.27961601151360405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,7168,0.20103466510772705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,10240,0.2544231149885389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,8192,0.2208355532752143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,6144,0.15846489535437688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,7168,0.06829600201712714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,8192,0.2046426667107476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,7168,0.2396479977501763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,7168,0.18284444014231363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,5120,0.12477066781785752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,6144,0.16575911309983996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,6144,0.05980177720387777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,51200,1.3588195376926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,4096,0.10547466410530938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,5120,0.05207733313242594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,6144,0.15919200579325357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,5120,0.13775822851392958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,3584,0.09024088912540013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,4096,0.04337066743108961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,4096,0.11107022232479519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,4096,0.11351466178894043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,3584,0.038070221741994224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,3072,0.07993244462543063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,3584,0.0985377762052748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,2560,0.06319110923343234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,3584,0.1116195519765218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,3072,0.03365688853793674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,3072,0.08683733145395915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,3072,0.08858311176300049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,2048,0.05186577637990316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,2560,0.07502310805850558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,2560,0.07703644699520536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,1536,0.040543999936845564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,2048,0.026410667432679072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,2048,0.06144977940453423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,2048,0.06617422236336602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,1024,0.028174221515655518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,1536,0.021728888154029846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,1536,0.04942044284608629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,5120,0.13481777244144016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,768,0.023580445183648005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,1024,0.018402665853500366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,1024,0.03802666730350918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,1024,0.042360001140170626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,8192,0.08224888642628987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,768,0.031154665682050917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,768,0.03578844335344102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,2560,0.030585779084099665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,512,0.015473778049151102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,256,0.012973333398501078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,512,0.024898666474554274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,128,0.009658666948477427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,512,0.031168888012568157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,256,0.014465777410401238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,256,0.02204711072974735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,64,0.00832622249921163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,128,0.014278221461508008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,128,0.021264889174037512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,32,0.01108266661564509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,128,0.018591110905011494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,1536,0.05452711052364773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,64,0.021288888321982488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,8192,32,0.021254221598307293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,8192,512,0.018438221679793466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,65536,0.5369226667616103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,8192,256,0.023434667123688593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,65536,1.3146675957573783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,51200,1.0358408821953666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,16384,0.3578488826751709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,8192,768,0.017027555240525138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,51200,0.4179848829905192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,16384,0.43300710784064395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,16384,0.12610577212439641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,12288,0.2756462097167969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,65536,2.0176853603786893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,51200,1.7653581831190321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,10240,0.2237475448184543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,12288,0.1144133276409573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,65536,1.5461680094401042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,12288,0.29298310809665257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,8192,0.1845066679848565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,12288,0.2703208923339844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,10240,0.26614401075575084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,51200,1.2353110843234592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,8192,0.06708977619806926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,7168,0.16309777895609537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,7168,0.061702224943372935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,8192,0.20959911081526014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,7168,0.18717688984341094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,6144,0.1438657840092977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,7168,0.16600799560546875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,6144,0.055989331669277616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,6144,0.151455110973782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,16384,0.3752133316463894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,6144,0.14650489224327937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,5120,0.04541777902179294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,10240,0.10039733515845405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,5120,0.12683644559648302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,5120,0.12397421730889215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,10240,0.2328044573465983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,8192,0.19276711675855848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,4096,0.09967199961344402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,4096,0.10398133595784505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,3584,0.03436177637841966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,3072,0.06931822167502509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,3584,0.0882106688287523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,3584,0.09142222007115682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,3072,0.07761333386103313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,5120,0.10734666718377008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,3072,0.08297866582870483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,4096,0.09574221902423435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,2560,0.07207377751668294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,2560,0.027961777316199407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,2048,0.0462097790506151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,3584,0.07859733369615343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,2048,0.02459200057718489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,1536,0.03574488891495599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,2048,0.061101330651177295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,3072,0.031151112582948472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,2560,0.05424177646636963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,1536,0.05051555567317539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,1024,0.026123555170165166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,1536,0.020849777592553034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,2560,0.0673875543806288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,1024,0.017571555243598092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,768,0.01977422171168857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,1024,0.03446222345034281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,1024,0.0395395557085673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,512,0.015023110641373528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,4096,0.03911022345225016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,768,0.015858666764365304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,768,0.028115553988350764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,256,0.01088444474670622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,768,0.03291022115283542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,512,0.014504000544548035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,512,0.022658665974934895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,128,0.010301333334710863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,512,0.028394665982988145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,256,0.019926221834288705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,256,0.021818665994538203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,1536,0.044608000251981944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,64,0.008728000024954477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,128,0.013785777820481194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,7168,32,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,7168,128,0.01756000022093455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,64,0.0192577772670322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,32,0.01924888955222236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,2048,0.05561421977149116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,7168,256,0.013533333109484779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,7168,128,0.019302222463819716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,51200,0.9220630857679578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,16384,0.32920887735154897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,65536,1.7083430820041234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,51200,0.3898933463626438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,16384,0.12589155303107366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,65536,1.3669840494791667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,16384,0.3482382297515869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,51200,1.343280898200141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,12288,0.2698524528079563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,16384,0.3316364553239611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,65536,1.259351094563802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,12288,0.09319911400477092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,51200,1.069336043463813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,12288,0.2671084403991699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,10240,0.20115910636054146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,8192,0.18168266614278158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,10240,0.07584800322850545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,12288,0.2456782129075792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,10240,0.20994222164154053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,7168,0.16374488671620688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,8192,0.0678115553326077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,8192,0.17280533578660753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,10240,0.20625600549909803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,8192,0.17874488565656874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,6144,0.13007467322879368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,7168,0.05500622259245979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,7168,0.14943110942840576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,5120,0.10974222421646118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,6144,0.047468443711598717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,7168,0.1477875577078925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,65536,0.4853990872701009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,6144,0.13015466266208225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,4096,0.08636266655392116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,5120,0.04173333446184794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,5120,0.10836533043119644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,3584,0.07253599829143949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,5120,0.11083022091123794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,4096,0.034824000464545354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,3584,0.03089955449104309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,3584,0.07999822166230944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,3584,0.08203288581636217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,3072,0.06776533524195354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,2560,0.05322844452328152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,3072,0.07367733452055189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,2560,0.05842489004135132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,2560,0.025337777203983728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,2560,0.06343377961052789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,2048,0.04900533292028639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,6144,0.13206666045718723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,2048,0.05450400047832065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,1536,0.03407999873161316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,2048,0.022037333912319605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,3072,0.05949422385957506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,1024,0.0248026665714052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,4096,0.09285600317849053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,1536,0.01827022267712487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,1536,0.038986666334999934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,1536,0.04559377829233805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,3072,0.028752889898088243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,768,0.020901333954599168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,1024,0.015607110328144498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,1024,0.03001333276430766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,512,0.01737244427204132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,1024,0.03597955571280585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,768,0.014467555615637036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,768,0.025366221865018208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,768,0.0303164455625746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,256,0.0131217779384719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,512,0.01346577786737018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,512,0.02000799940692054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,128,0.011531555818186866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,512,0.025303110480308533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,256,0.012692444854312472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,256,0.017924444542990792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,64,0.010460444622569613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,256,0.021095999413066443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,6144,128,0.012459555433856117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,32,0.010793777803579966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,128,0.017197334104114108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,4096,0.0869564414024353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,64,0.01723022262255351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,6144,32,0.01719999975628323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,6144,2048,0.041950222518708974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,65536,0.43186132113138836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,65536,1.1590799755520291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,6144,128,0.017139555679427255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,65536,1.4403564665052626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,51200,0.35263821813795304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,16384,0.27426666683620876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,16384,0.098907556798723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,51200,1.1783822377522786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,16384,0.2981831232706706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,12288,0.2114888827006022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,16384,0.29657867219713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,51200,0.7744480239020454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,51200,0.9156906339857313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,12288,0.0773635572857327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,10240,0.17314044634501138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,12288,0.21429599655999076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,8192,0.12596089310116237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,10240,0.07221333185831706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,12288,0.23443465762668184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,10240,0.18422133392757842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,7168,0.12227733929951985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,8192,0.052926222483317055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,8192,0.1622435516781277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,10240,0.18126844035254586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,6144,0.10026933087242974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,8192,0.14964000384012857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,7168,0.04974311259057787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,7168,0.130566226111518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,5120,0.07450844181908502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,6144,0.04131644301944309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,7168,0.1331191062927246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,65536,1.2250693639119465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,6144,0.10991022321912979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,4096,0.06724266873465644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,5120,0.035827553934521146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,6144,0.11533066961500381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,3584,0.05159644285837809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,5120,0.0950311091211107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,4096,0.030748443471060857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,5120,0.0989111132091946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,3072,0.04978222317165799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,4096,0.07578844494289823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,3584,0.02742133372359806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,2560,0.03894755575391982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,3584,0.06713777780532837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,3072,0.025115556187099878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,3584,0.07379111316468981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,2048,0.03286577926741706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,3072,0.05984977881113688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,2560,0.022892443670166865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,3072,0.06605599986182319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,2560,0.051049778858820595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,1536,0.02600266701645321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,2048,0.019719999697473314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,2560,0.05752355522579617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,1024,0.019009777241282992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,2048,0.042698668109046094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,1536,0.017445334129863314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,768,0.01548266741964552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,1536,0.03416088885731167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,1024,0.014835556348164877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,1536,0.04023111197683547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,512,0.011556444068749746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,1024,0.03257511059443156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,768,0.013779555757840475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,768,0.022408889399634466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,256,0.008720888859695857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,768,0.027076444692081873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,512,0.017771555317772757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,512,0.023010666171709698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,4096,0.08217422167460124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,128,0.0075795559419526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,256,0.01591377788119846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,256,0.01959911154376136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,128,0.011776888536082374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,64,0.008292444050312042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,2048,0.04956533180342781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,5120,32,0.00851022203763326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,64,0.01519377695189582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,1024,0.026224000586403742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,32,0.015202666322390238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,512,0.0125591109196345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,5120,256,0.011709333293967776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,65536,0.36452354325188535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,5120,128,0.015830222103330824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,65536,1.0119768778483074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,16384,0.2716968854268392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,65536,1.2266764110989041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,51200,0.27872443199157715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,16384,0.08362222380108303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,65536,1.020536846584744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,16384,0.2557786570654975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,12288,0.19425777594248453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,16384,0.2597422334882948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,51200,0.9459315405951606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,51200,0.7710586653815376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,12288,0.18611999352773032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,12288,0.06495910882949829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,5120,128,0.01513066722287072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,10240,0.16130933496687147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,8192,0.12830132908291286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,10240,0.05712178018358019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,10240,0.15962221887376574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,7168,0.1068391137652927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,8192,0.12005689409044053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,8192,0.044686221414142184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,10240,0.16537422604031032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,8192,0.13409688737657335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,6144,0.09102400143941243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,7168,0.039847999811172485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,7168,0.1100719968477885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,6144,0.03561688794030084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,5120,0.07032622231377496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,6144,0.09461688995361328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,7168,0.1171848906411065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,6144,0.10264888736936782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,4096,0.05596888727611966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,5120,0.030702223380406696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,5120,0.07816355758243136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,3584,0.05215377940071953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,5120,0.0874204436937968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,4096,0.02645066711637709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,4096,0.0641288889778985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,3072,0.04243288768662346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,4096,0.07386133405897352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,3584,0.02418222195572323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,12288,0.19561066892411974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,3584,0.06518933508131239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,2560,0.035183111826578774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,3072,0.05059733324580722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,2048,0.028130667077170476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,51200,0.6848666403028699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,2560,0.019718221492237516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,2560,0.04363110992643568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,3072,0.057681779066721596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,1536,0.023393778337372675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,2048,0.01735466718673706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,2560,0.05067200130886502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,2048,0.036730666955312095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,1024,0.017055999901559617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,2048,0.04379466507169935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,1536,0.014866666661368476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,1536,0.029659556017981634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,768,0.013394667042626275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,1024,0.013713777893119387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,1536,0.03566133313708835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,512,0.010625777973069085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,1024,0.02792000108295017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,768,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,768,0.019512888458040025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,768,0.02443555494149526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,256,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,512,0.011615999870830111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,512,0.01553066737122006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,3584,0.05681600173314413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,128,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,512,0.0207324441936281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,256,0.010839111275143094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,64,0.007483555210961237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,3072,0.02218666672706604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,256,0.01812799937195248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,4096,32,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,4096,128,0.010984000232484607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,128,0.01310400002532535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,4096,128,0.014418666561444601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,64,0.013108444710572561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,32,0.01315199997689989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,1024,0.022273777259720698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,65536,0.3068026701609294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,65536,0.7008426454332141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,51200,0.4586888949076335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,16384,0.1831617752710978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,51200,0.24061955346001518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,65536,1.0063003963894315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,4096,256,0.013790222505728403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,16384,0.2219555510414971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,51200,0.7837360170152453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,12288,0.1404728889465332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,16384,0.0748951103952196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,16384,0.2656702200571696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,10240,0.12405955791473389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,12288,0.056594669818878174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,51200,0.6829635302225748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,12288,0.16066843933529323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,8192,0.09721333450741237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,12288,0.16812711291842988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,10240,0.049795554743872754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,10240,0.13646222485436335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,7168,0.08868444628185695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,8192,0.0396106673611535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,8192,0.10920355717341106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,10240,0.139082670211792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,6144,0.06615910927454631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,7168,0.03592177894380357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,8192,0.12057067288292779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,7168,0.09655644496281941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,5120,0.05551911062664456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,7168,0.10238755411571926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,6144,0.031630221340391375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,6144,0.08342133627997504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,4096,0.044547554519441396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,6144,0.08828088972303604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,65536,0.851835568745931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,5120,0.07067644596099854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,3584,0.03834133346875509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,5120,0.07727377944522433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,4096,0.05792178048027886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,4096,0.023504889673656885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,3072,0.03327111072010464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,4096,0.06416089004940458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,3584,0.02160266704029507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,3584,0.052296890152825244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,2560,0.02847377790345086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,3584,0.05730666716893514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,3072,0.04632800155215793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,3072,0.020032889313167997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,2048,0.02327199942535824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,3072,0.0511084430747562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,2560,0.017873777283562552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,2560,0.04019200139575534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,1536,0.018790221876568265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,2560,0.04558488726615906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,2048,0.03388355506791009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,2048,0.01575199928548601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,1024,0.014125333891974555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,2048,0.038633777035607234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,1536,0.01419911119672987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,768,0.011039999624093374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,1536,0.02695111102528042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,1536,0.03238577644030253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,1024,0.012318222059143914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,1024,0.02074399921629164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,512,0.008803555534945594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,5120,0.02784444557295905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,768,0.011374221907721626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,256,0.0063324446479479475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,1024,0.025239111648665533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,768,0.017641777793566387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,128,0.00618488879667388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,512,0.010710222025712332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,768,0.021911111142900255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,256,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,64,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,512,0.018918222851223417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,256,0.012735111018021902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3584,32,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,256,0.01666933298110962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,128,0.012422222230169507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3584,128,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,64,0.012287110918098025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,32,0.012063999970753988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,65536,0.5770862367418077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,51200,0.45452266269259983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,65536,0.8850240177578397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3584,512,0.014323555760913424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,51200,0.6848097907172309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3584,128,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,65536,0.9081911510891385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,16384,0.07031822204589844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,12288,0.13445244895087347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,16384,0.19017333454555937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,51200,0.6668257713317871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,16384,0.215265777375963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,10240,0.11350044939253066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,12288,0.054754667811923556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,65536,0.2880115509033203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,12288,0.14097332954406738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,12288,0.16893243789672852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,10240,0.047544889979892306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,10240,0.11986755000220405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,7168,0.07860800292756823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,8192,0.037273777855767146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,10240,0.1449928945965237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,8192,0.09647466739018758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,51200,0.22509778870476616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,8192,0.1145342191060384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,6144,0.06466044320000543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,7168,0.033879111210505165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,16384,0.16767999860975477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,7168,0.08564888768725926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,5120,0.05540977915128072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,7168,0.09910221894582112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,6144,0.03022577696376377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,4096,0.04596711198488871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,6144,0.07448889149559869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,5120,0.026475555366939966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,6144,0.08697955475913154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,5120,0.06310400035646227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,3584,0.03794577717781067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,5120,0.07567466629876031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,4096,0.0225582222143809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,4096,0.05110044280687968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,4096,0.06322044134140015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,3584,0.02064533366097344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,8192,0.08921333154042561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,3584,0.04627466532919142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,2560,0.028949333561791316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,3584,0.05643733342488607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,3072,0.0189057770702574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,3072,0.04051288962364197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,2048,0.02474044428931342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,2560,0.017349332571029663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,3072,0.05054310957590739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,1536,0.01977599991692437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,2560,0.044052445226245456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,2048,0.01552177800072564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,2048,0.02943555514017741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,1024,0.014075555735164218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,2048,0.03795555565092299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,1536,0.014020444618331062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,1536,0.0240639994541804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,768,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,1536,0.0314648879898919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,1024,0.012073777616024017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,1024,0.018384888768196106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,512,0.0087013335691558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,1024,0.025087111526065405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,768,0.011427555647161273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,768,0.016125332978036668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,3072,0.03317688902219137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,256,0.006562666760550604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,768,0.02127200033929613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,512,0.01291822228166792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,128,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,512,0.018633777896563213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,256,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,256,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,64,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,3072,32,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,2560,0.03565333286921183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,128,0.009995555712117089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,128,0.011232888533009423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,128,0.012412444584899478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,64,0.011151110960377587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,3072,32,0.011047110789351992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,65536,0.5277706782023112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,65536,0.2561431196000841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,51200,0.36572265625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,16384,0.12980000178019205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,3072,512,0.010727110836240979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,51200,0.2050080034467909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,65536,0.6931084526909722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,51200,0.5888942082722982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,3072,256,0.01630577776167128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,16384,0.06490666336483426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,12288,0.10884355836444431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,16384,0.18228444788191053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,10240,0.09259288840823704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,12288,0.04974666568968031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,16384,0.17872711022694907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,51200,0.5568488968743218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,8192,0.06835911009046766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,10240,0.043177776866488986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,12288,0.13792622089385986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,7168,0.0627982219060262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,10240,0.12475644217597114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,8192,0.034670223792394005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,8192,0.08569422033098008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,6144,0.049623999330732554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,8192,0.09517155753241645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,7168,0.0313662224345737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,7168,0.07514933082792494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,65536,0.7305582364400228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,5120,0.044010665681627065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,7168,0.08432000213199192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,6144,0.028338667419221666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,6144,0.0652186671892802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,4096,0.03609422180387709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,6144,0.07442666424645318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,5120,0.023807111713621352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,3584,0.032130665249294706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,5120,0.05510044429037306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,5120,0.06451555755403307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,4096,0.045590221881866455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,4096,0.05474844574928284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,12288,0.12492088476816814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,3072,0.0259226659933726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,3584,0.01826400061448415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,3584,0.04112799962361654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,10240,0.10455822282367283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,2560,0.02401955591307746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,3584,0.048335111803478666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,3072,0.0169493340783649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,3072,0.03580266568395827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,2048,0.01868799991077847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,2560,0.015483554866578845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,3072,0.04360533422893948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,2560,0.03144355614980062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,1536,0.015257777439223396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,2560,0.03784711162249247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,2048,0.015060444672902426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,1024,0.010947555303573608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,2048,0.026029333472251892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,2048,0.03321244319279989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,1536,0.012468444804350534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,1536,0.021420444051424663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,1536,0.02734044525358412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,1024,0.010970667004585266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,1024,0.01667022208372752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,512,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,1024,0.021985777550273474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,768,0.010339555641015371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,768,0.014448889427714877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,4096,0.02090399960676829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,256,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,768,0.01902933253182305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,512,0.009704888694816166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,512,0.011761777930789523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,128,0.005072000126043956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,512,0.017920888132519193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,256,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,64,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,256,0.015078221758206686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,128,0.0094355551732911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,32,0.004523555437723796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,128,0.010130667024188572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2560,128,0.012063999970753988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,64,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2560,32,0.010032888915803697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2560,768,0.008834666675991481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,65536,0.21779466999901664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,51200,0.3416782220204671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,65536,0.5748577647738987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,16384,0.12382045057084824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,51200,0.17162756125132242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,65536,0.6776257620917426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2560,256,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,51200,0.4587066438462999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,16384,0.14116444852617052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,16384,0.05734311209784614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,12288,0.09041600094901191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,10240,0.07742755280600654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,16384,0.1768702268600464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,12288,0.04480266571044922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,51200,0.5289857652452257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,12288,0.10644888877868652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,8192,0.06338222159279717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,10240,0.03869333200984531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,12288,0.1354773309495714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,10240,0.09171289205551147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,7168,0.05480088790257772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,65536,0.4364648924933539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,8192,0.030301332473754883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,10240,0.11470488707224528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,8192,0.0735093355178833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,7168,0.02754577828778161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,8192,0.0943031112353007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,5120,0.03862844573126899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,7168,0.06486133072111341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,7168,0.08337066570917766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,6144,0.024538666009902954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,6144,0.055758224593268506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,4096,0.028749333487616643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,5120,0.021418665846188862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,3584,0.0269128895468182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,5120,0.04730133215586344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,4096,0.018650665879249573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,5120,0.06299110915925768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,4096,0.039157334301206805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,3072,0.022431110342343647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,3584,0.017044444878896076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,3584,0.03538044293721517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,2560,0.02001422146956126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,3072,0.015864888827006023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,3584,0.04807555675506592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,3072,0.03160711129506429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,3072,0.04244177871280246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,2560,0.014835556348164877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,2560,0.026943110757403906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,2560,0.0374017788304223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,6144,0.044305778212017484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,2048,0.023244443866941664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,1536,0.013079110946920184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,2048,0.013447999954223633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,2048,0.03163822160826789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,1536,0.012147555748621622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,1024,0.009515555368529426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,1536,0.018935110833909776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,1536,0.026416889495319788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,6144,0.07333688603507148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,768,0.007736889024575551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,1024,0.010660444696744284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,1024,0.0148044443792767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,512,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,4096,0.053120000494851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,768,0.010001777774757808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,768,0.012700444294346703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,256,0.004945777770545748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,768,0.018548443913459778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,512,0.009324444664849175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,512,0.010486221975750394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,2048,0.0162453336848153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,128,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,512,0.016177778442700703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,256,0.009032888544930352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,256,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,64,0.004535111288229625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,2048,32,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,256,0.014879110786649915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,2048,128,0.009030222064918941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,128,0.009169778062237633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,128,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,64,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,2048,32,0.009353777600659264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,51200,0.19891288545396593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,65536,0.17238756020863852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,2048,1024,0.021528000632921856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,65536,0.46915377510918516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,16384,0.07996266418033175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,51200,0.13629155688815647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,51200,0.368468443552653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,65536,0.5330035421583388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,12288,0.06826755735609266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,16384,0.046499556965298124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,16384,0.11895999643537734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,10240,0.05518844392564562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,51200,0.4182257652282715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,12288,0.036107556687461004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,16384,0.14303555753495958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,12288,0.09017688698238796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,8192,0.0409733321931627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,12288,0.11031999852922227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,10240,0.03178400132391188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,7168,0.03833599885304769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,10240,0.07472533649868436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,10240,0.09312888648774888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,8192,0.061478224065568715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,6144,0.031155556440353394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,8192,0.07650399870342679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,7168,0.05456533365779453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,5120,0.026399110754330952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,7168,0.06799377997716267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,6144,0.0203128887547387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,6144,0.047444442907969155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,4096,0.02181244393189748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,6144,0.060418665409088135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,5120,0.018018666240904067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,5120,0.04045155644416809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,3584,0.019277332557572257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,4096,0.015843555331230164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,4096,0.03360355562633938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,4096,0.04423288835419548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,3072,0.016566221912701923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,3584,0.014656000667148165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,3584,0.0303111109468672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,3584,0.039213332864973284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,2560,0.014460444450378418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,65536,0.26887467172410756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,3072,0.02678844498263465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,3072,0.03510399990611606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,7168,0.022271111607551575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,2048,0.012062222593360476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,2560,0.02379911144574483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,2560,0.031110223796632554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,2048,0.011401777466138205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,2048,0.019944889677895438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,2048,0.027191110783153113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,5120,0.05188266767395867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,1536,0.016813332835833233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,1536,0.01035288886891471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,1024,0.00792711145348019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,1536,0.022447110878096685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,768,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,1024,0.009677333136399588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,1024,0.013440888788965015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,1024,0.01841600073708428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,512,0.005783111270931032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,768,0.011195555329322815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,8192,0.02481422159406874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,768,0.009300444689061906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,256,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,768,0.016504888733228047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,512,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,2560,0.01259022206068039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,512,0.009663110805882348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,128,0.0038622220357259116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,512,0.014827556080288358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,256,0.00850311087237464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,256,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,256,0.013080000049538083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,128,0.008273777862389883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,32,0.0038355555799272326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,128,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1536,128,0.010415111151006486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,64,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1536,32,0.008282666405042013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,65536,0.16443110836876765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1536,3072,0.013678222066826291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,65536,0.3733866744571262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,51200,0.19371644655863443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,65536,0.5284533500671387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,1536,0.01035911093155543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,51200,0.13075021902720133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,51200,0.29351377487182617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,16384,0.044235554006364614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,51200,0.4140453338623047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,16384,0.09733244445588853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,12288,0.047672887643178306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,65536,0.23402044508192274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,16384,0.1422542201148139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,10240,0.03700088792377048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,12288,0.03403910994529724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,12288,0.07287822167078654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,10240,0.03037244412634108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,8192,0.028637333048714533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,12288,0.10905777745776707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,10240,0.06127999888526069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,7168,0.0265964451763365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,10240,0.09058488739861383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,8192,0.023214222656355962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,8192,0.05007733239067925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,7168,0.021363554729355708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,7168,0.0443173348903656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1536,64,0.003546666767862108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,16384,0.06372266345553927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,7168,0.06716177860895793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,5120,0.019519110520680744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,6144,0.018944889307022095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,6144,0.03920888900756836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,4096,0.016504888733228047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,5120,0.0170462214284473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,6144,0.05910578038957384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,5120,0.03329777717590332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,3584,0.014540443817774454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,5120,0.05091466506322225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,4096,0.015126221709781222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,3072,0.0129466669427024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,4096,0.027906666199366253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,3584,0.014035556051466199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,4096,0.04218222366438972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,3584,0.025109334124459162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,2560,0.011064888702498542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,3584,0.03825689024395413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,3072,0.02239822182390425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,3072,0.01275199982855055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,2048,0.009005332986513773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,3072,0.03348533312479655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,2560,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,2560,0.019708444674809773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,1536,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,2560,0.02940088841650221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,2048,0.01106844428512785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,6144,0.023257777094841003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,2048,0.016884444488419425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,1024,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,2048,0.025910221868091162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,1536,0.01000533335738712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,768,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,1536,0.014130666851997375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,1024,0.00928177767329746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,512,0.004522666749027041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,1024,0.011224889092975192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,768,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,1024,0.018204443984561496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,768,0.009764444496896531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,768,0.015962666935390897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,512,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,512,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,128,0.003170666595300039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,512,0.014343111051453484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,256,0.00833511104186376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,256,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,64,0.003551111039188173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,128,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,1024,128,0.008044444852405125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,32,0.0038293335172865125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,128,0.010061333576838175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,64,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,65536,0.17924533949957955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,1024,32,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,51200,0.14103644424014622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,65536,0.16183555126190186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,1536,0.021987555755509272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,65536,0.32733334435356987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,1024,256,0.0038213332494099936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,65536,0.38413688871595597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,51200,0.12869866689046225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,51200,0.2564320034450955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,51200,0.305369774500529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,16384,0.04377955529424879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,256,0.01296533313062456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,16384,0.08565511306126912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,12288,0.03625244564480252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,10240,0.029186666011810303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,16384,0.10701421896616618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,12288,0.033760001262029014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,12288,0.06436889039145575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,1024,8192,0.0755573312441508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,8192,0.023710222707854375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,10240,0.030004445049497817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,10240,0.05482133229573568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,12288,0.08181244134902954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,7168,0.02070844504568312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,10240,0.0692195561197069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,8192,0.023030221462249756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,8192,0.044637332359949745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,7168,0.020655110478401184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,7168,0.03987377882003784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,16384,0.04788000053829617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,5120,0.016216889023780823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,6144,0.018933332628673978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,6144,0.03523911039034525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,6144,0.04551911022928026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,5120,0.016914667354689706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,4096,0.013551111022631327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,5120,0.029912001556820337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,5120,0.03951466745800442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,3584,0.012142221960756512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,4096,0.014855111638704935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,4096,0.025032889511850145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,4096,0.03303111261791653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,3072,0.01053244454993142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,3584,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,3584,0.0229066660006841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,2560,0.009593778186374241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,3584,0.029457777738571167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,3072,0.012825777961148156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,6144,0.018852444158660043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,3072,0.020344000723626878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,8192,0.057877335283491343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,3072,0.026721777187453374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,2560,0.012049777640236748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,2560,0.018030222919252183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,1536,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,2560,0.023757333556811016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,2048,0.01108000013563368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,2048,0.015440889530711703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,1024,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,2048,0.020959110723601446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,1536,0.012846222354306115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,1536,0.010061333576838175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,768,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,1536,0.018233777748213876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,1024,0.009339555270142024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,1024,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,512,0.003866666720973121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,768,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,768,0.009148444566461775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,256,0.0032257777121331957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,768,0.013762666947311826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,512,0.008550222549173567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,512,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,128,0.0028906667398081887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,512,0.01274400038851632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,256,0.008120888637171851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,256,0.0069919998447100324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,64,0.002830222249031067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,256,0.011351111034552256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,2048,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,768,128,0.008005333443482717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,768,32,0.0028328889360030494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,128,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,7168,0.05145155390103658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,65536,0.15230488777160645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,64,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,768,32,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,51200,0.11288532945844863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,65536,0.11812978320651585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,65536,0.28443556361728245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,16384,0.03898933198716905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,51200,0.09885866774453057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,51200,0.22388267517089844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,1024,0.015250666273964776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,65536,0.38172266218397355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,12288,0.026983110441101923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,16384,0.03398222062322829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,16384,0.07430044147703382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,51200,0.2983893288506402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,10240,0.02260444396071964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,12288,0.024849777420361836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,16384,0.1057422227329678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,8192,0.01830044388771057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,12288,0.05695911248524984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,10240,0.02330400049686432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,12288,0.08048977454503377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,10240,0.0480000012450748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,7168,0.01610577768749661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,8192,0.017496888836224873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,10240,0.06851644648445977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,8192,0.03957155678007338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,6144,0.014702222413486905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,768,128,0.008984888593355814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,8192,0.05674488676918877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,7168,0.016022221909628976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,5120,0.012673777838548025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,7168,0.03546844588385688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,6144,0.014490667316648694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,6144,0.030972444348865088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,7168,0.05094844433996412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,4096,0.010592888626787396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,5120,0.01312355531586541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,6144,0.04490400022930569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,5120,0.026595556073718604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,3584,0.00958933350112703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,4096,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,5120,0.03847200009557936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,3072,0.00868711123863856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,4096,0.022267555197079975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,3584,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,4096,0.032066666417651706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,2560,0.007608888877762689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,3072,0.010501333408885531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,3584,0.029205332199732464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,3072,0.018234666850831773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,2048,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,3072,0.02647288971477085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,2560,0.010037333601050908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,1536,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,2560,0.023429332507981196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,2048,0.009478222164842818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,2048,0.013815999858909182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,1024,0.004516444272465176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,2048,0.020953777763578627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,1536,0.009007111191749573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,1536,0.018138666947682697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,1024,0.008367111285527548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,1024,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,1024,0.015052444405025907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,768,0.008007110820876228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,768,0.008370666868156856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,512,0.0034933334423436057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,768,0.01312355531586541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,3584,0.0203137778573566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,256,0.002883555574549569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,512,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,512,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,512,0.012420444852775998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,128,0.0025822222232818604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,2560,0.015916445189052157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,256,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,256,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,64,0.002471999989615546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,32,0.002536888958679305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,512,128,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,256,0.010716444088353051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,128,0.006144000010357962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,1536,0.011630222201347351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,65536,0.07921866575876872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,512,128,0.008428444465001425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,64,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,51200,0.0636319981680976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,512,768,0.0038942222793896994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,16384,0.02502222193611993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,65536,0.24383287959628633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,51200,0.09381066428290473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,51200,0.19154844019148085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,12288,0.019309333629078336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,16384,0.03247377938694424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,65536,0.365372445848253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,16384,0.06373511420355903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,51200,0.2883831130133735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,10240,0.018908444378111098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,16384,0.10180622339248657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,12288,0.04855377806557549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,8192,0.016557332542207506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,10240,0.04124622212515937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,10240,0.022392888863881428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,12288,0.07719466421339247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,7168,0.01389422184891171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,10240,0.06555822160508898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,8192,0.016885333591037326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,8192,0.03408177693684896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,6144,0.012978666358523898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,512,32,0.005969777703285217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,7168,0.030288888348473444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,7168,0.048186666435665555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,65536,0.11699644724527995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,6144,0.014357333381970724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,6144,0.026757333013746474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,6144,0.04241066508822971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,4096,0.009256000320116678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,5120,0.013057777451144325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,5120,0.035903112755881414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,12288,0.02472977836926778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,4096,0.01959466603067186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,3584,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,4096,0.011711111499203576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,4096,0.030792888667848375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,3584,0.011063999599880643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,3072,0.006567111031876669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,3584,0.017844445175594754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,7168,0.015531554818153381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,8192,0.054401778512530856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,2560,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,5120,0.010429333481523724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,3072,0.010418666733635796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,3072,0.016075554821226332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,2048,0.004864888886610667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,3072,0.025126222107145522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,2560,0.01001422190003925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,2560,0.014261333478821648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,1536,0.004566222015354368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,5120,0.02289955483542548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,2048,0.009538667069541084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,2048,0.0123857781291008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,1024,0.003551111039188173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,2048,0.019850666324297588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,1536,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,1536,0.010421333213647207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,768,0.0032106666929192017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,1536,0.01721244388156467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,1024,0.008292444050312042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,1024,0.014429334137174817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,512,0.0028728888266616394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,768,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,768,0.007772444850868649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,768,0.012712000144852532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,3584,0.027895109521018133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,256,0.002467555511328909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,512,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,512,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,128,0.002247111043996281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,512,0.011695110963450538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,256,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,256,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,256,0.010363555616802638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,256,128,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,32,0.00221422231859631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,2560,0.022637334134843614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,256,128,0.008026666939258575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,65536,0.059161775641971164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,32,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,51200,0.04907911022504171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,1024,0.00867022242810991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,65536,0.11547822422451443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,65536,0.2053911156124539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,16384,0.015729778342776828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,51200,0.09222400188446045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,65536,0.3625920083787706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,51200,0.16089067194196913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,12288,0.011997333003415002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,16384,0.05433244506518046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,16384,0.03196355700492859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,51200,0.28640709982977974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,10240,0.011046222514576383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,256,64,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,12288,0.02404622236887614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,12288,0.041631112496058144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,8192,0.010047111246320937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,10240,0.02179466684659322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,12288,0.07670044236712985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,10240,0.03511555658446418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,7168,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,10240,0.06514044602711995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,8192,0.01711466742886437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,6144,0.009022222624884712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,8192,0.053210665782292686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,7168,0.015573332707087198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,7168,0.0258942229880227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,5120,0.008777778181764815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,6144,0.014347554908858405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,6144,0.023099554909600153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,6144,0.04166488846143087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,4096,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,5120,0.01308444473478529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,5120,0.020015110572179157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,5120,0.03575911124547323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,4096,0.011535111400816174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,3584,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,16384,0.1005644467141893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,4096,0.030493332280053034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,3072,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,3584,0.011095110740926532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,3584,0.015829333000712924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,256,64,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,2560,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,3584,0.02777155571513706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,8192,0.029072887367672388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,3072,0.010629333555698395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,3072,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,2048,0.004855999930037392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,2560,0.010037333601050908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,7168,0.04788622260093689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,2560,0.012759110993809171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,1536,0.0038986665507157645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,2560,0.022325333621766832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,2048,0.011229332950380115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,2048,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,2048,0.019880889190567862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,1024,0.0032702222880389956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,1536,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,1536,0.009650666680600908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,768,0.00315022220214208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,1536,0.017216889394654166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,1024,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,4096,0.01719199948840671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,1024,0.01440711153878106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,512,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,768,0.00797155582242542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,768,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,256,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,768,0.012695111334323883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,512,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,512,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,128,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,512,0.011799111134476133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,256,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,256,0.010051555931568146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,64,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2560,128,128,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,128,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,128,32,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,128,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,65536,0.05731733640034994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,51200,0.046260442998674184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,64,0.0052888890107472735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,32,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,16384,0.01848266687658098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,12288,0.010792888700962067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,1024,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,65536,0.20416888925764295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,10240,0.00980799976322386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,8192,0.008630221916569604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,51200,0.16027911504109701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,16384,0.054216000768873424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,7168,0.010681777364677854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,12288,0.04122044311629401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,6144,0.009875555833180746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,5120,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,10240,0.03478844298256768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,8192,0.028537776735093858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,4096,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,3584,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,7168,0.025751999682850305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,3072,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,128,256,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,6144,0.023046221998002794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,2560,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,5120,0.019976889093716938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,2048,0.004556444370084339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2560,128,3072,0.02499733368555705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,1536,0.0038453332251972626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,1024,0.0031742221779293488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,768,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,4096,0.017100445098347135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,3584,0.01569955547650655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,512,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,256,0.0022684445397721394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,2048,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,1536,0.009639110830095079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,128,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,1024,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,64,0.001957333336273829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,64,32,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,512,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,256,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,65536,0.05728799766964383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,128,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,64,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,51200,0.04483200112978617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,32,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,12288,0.009880888793203566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,65536,0.2045066621568468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,16384,0.054195556375715465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,51200,0.1605102221171061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,10240,0.008961777720186446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,8192,0.008026666939258575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,12288,0.041264888313081526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,7168,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,6144,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,10240,0.03490044342146979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,8192,0.028616888655556574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,7168,0.025769778423839148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,3072,0.014144000079896716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,6144,0.022985777921146814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,4096,0.007879111501905654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,3584,0.007132444116804335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,3072,0.006322666588756774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,768,0.007339555356237624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,5120,0.01993688941001892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,2560,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,2048,0.004555555681387584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,4096,0.017045333981513977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,3584,0.015560888581805758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,1536,0.0038355555799272326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,3072,0.014172444740931192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,1024,0.0032008888406885993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,16384,0.012392000191741519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,768,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,2560,0.012763555679056378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,512,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,2048,0.011193777951929303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,256,0.0022124445272816554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,1536,0.009512888888518015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,128,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,1024,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,64,0.0018826667219400406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,768,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,512,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,256,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,128,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,64,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,5120,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,32,32,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2560,64,2560,0.012660444610648684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2560,32,32,0.0018639999131361644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,16384,1.0081279542711046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,12288,0.7755751079983182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,16384,3.2231012980143228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,12288,2.5653101603190103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,16384,3.2923492855495877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,12288,2.343382305569119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,10240,0.6183448897467719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,51200,3.6328790452745228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,10240,2.0438817342122397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,8192,1.5493591096666124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,8192,1.4878888660007055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,8192,0.4955057568020291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,10240,1.9106551276312933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,7168,0.4451706674363878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,6144,0.37934666209750706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,6144,1.0817093319363065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,7168,1.490559154086643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,6144,1.2285110685560439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,7168,1.3162462446424696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,5120,0.9986471600002713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,5120,0.328512880537245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,4096,0.27824534310234916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,4096,0.9049626456366645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,3584,0.6801964441935221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,3584,0.245399104224311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,3584,0.6726897557576498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,3072,0.21538489394717744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,3072,0.6448773278130425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,2560,0.4200133217705621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,3072,0.599455992380778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,5120,0.9630319807264539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,2560,0.19244000646803114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,2048,0.3641528818342421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,2560,0.47752798928154844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,1536,0.29134400685628253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,2048,0.157169779141744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,4096,0.732692453596327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,1536,0.12849333551194933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,2048,0.39878222677442765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,1024,0.229272895389133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,1024,0.10071288877063328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,768,0.17717422379387748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,1536,0.33224354849921334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,768,0.09515644444359674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,512,0.14485155211554632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,1024,0.22925954394870332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,256,0.08975378010008071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,512,0.08204266760084364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,768,0.18265510929955375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,256,0.07301866345935397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,512,0.16206933392418757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,256,0.140937778684828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,65536,128,0.06982044378916423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,32,0.05852178070280287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,51200,10.28281995985243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,128,0.13597333431243896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,64,0.13692000177171496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,32,0.1351742214626736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,128,0.07193422317504883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,65536,64,0.05774488714006212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,16384,2.403552797105577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,65536,51200,13.827888488769531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,51200,2.8354434967041016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,65536,3.6554586622450085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,16384,0.7746657795376248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,12288,0.6603795687357584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,16384,2.5777155558268228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,12288,2.0500852796766496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,12288,1.9371467166476781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,51200,7.444088406032986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,10240,1.5534231397840712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,10240,0.4959893226623535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,10240,1.5554702546861436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,8192,0.40892354647318524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,8192,1.3451733059353297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,7168,0.34270132912529844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,7168,1.271972444322374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,8192,1.1980533599853516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,7168,1.0034898122151692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,65536,9.629725138346354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,6144,0.9969760047064887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,6144,0.3118204540676541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,5120,0.752928892771403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,5120,0.25783199734157985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,6144,0.9161262512207031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,4096,0.7217573589748807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,4096,0.21507022115919325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,51200,10.49624294704861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,3584,0.5967653592427572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,5120,0.7324310938517252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,3584,0.19121955500708687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,4096,0.5837182468838161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,3072,0.4852995342678494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,3072,0.19230577680799696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,2560,0.4395297898186578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,2560,0.14649066660139295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,3584,0.6550675498114692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,3072,0.42970577875773114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,2048,0.3380177815755208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,2560,0.381386677424113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,1536,0.2569182184007433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,2048,0.14292532867855495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,1536,0.10024444262186687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,1024,0.1780631144841512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,768,0.1519200007120768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,2048,0.4099688794877794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,1024,0.18167466587490508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,1024,0.0810755557484097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,768,0.07946933640374078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,512,0.11369955539703369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,256,0.06688888867696126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,768,0.14422844515906438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,128,0.052016887399885386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,256,0.0710480014483134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,512,0.1265546613269382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,64,0.058881779511769615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,256,0.11468088626861572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,51200,32,0.05418133404519823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,128,0.10582755671607123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,64,0.10690044694476658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,32,0.10741066932678223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,1536,0.25086045265197754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,512,0.0655591090520223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,51200,128,0.05679821968078613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,51200,65536,13.624393039279513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,65536,1.1439217461480033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,16384,0.7998035748799642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,51200,0.8622204462687174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,16384,0.26078666581047905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,16384,0.8176133367750379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,51200,2.4063538445366754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,12288,0.19762399461534289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,12288,0.7247253523932563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,10240,0.45284446080525714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,10240,0.16439021958245173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,12288,0.599365340338813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,65536,3.242039998372396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,8192,0.13387644290924072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,7168,0.3681742350260417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,8192,0.407608879937066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,10240,0.4836622344122992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,7168,0.11925600634680854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,6144,0.32667554749382866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,7168,0.3670809004041884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,6144,0.10615822341707017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,5120,0.22447289360894096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,51200,3.5316026475694446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,4096,0.19775466124216715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,6144,0.2897813320159912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,65536,4.332305060492621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,3584,0.15739822387695312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,4096,0.07996355825000338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,5120,0.28670578532748753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,4096,0.195722672674391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,8192,0.40840533044603133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,3072,0.12944977813296848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,3584,0.06952177815967135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,2560,0.11592711342705621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,3072,0.06519110997517903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,3584,0.16805156071980795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,2560,0.05315555466545952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,3072,0.1510897742377387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,1536,0.08025066720114814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,2048,0.04588355620702108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,2560,0.12671821647220188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,2048,0.10435466633902656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,1536,0.037560890118281044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,768,0.04309422108862135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,1024,0.030981332063674927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,512,0.03071911136309306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,768,0.029081778393851385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,768,0.051651553975211255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,256,0.022710222336981032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,512,0.026523555318514507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,512,0.043819556633631386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,128,0.018272888329293992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,256,0.03978133201599121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,2048,0.09637688928180271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,128,0.03764444589614868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,128,0.023200000325838726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,64,0.015814221567577787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,32,0.01639111174477471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,1536,0.08577511045667861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,1024,0.06456533405515882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,32,0.03732088870472378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,5120,0.09089510970645481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,16384,256,0.024175999893082514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,16384,1024,0.051473776499430336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,65536,0.8625600073072645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,16384,64,0.03769866625467936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,51200,0.6929226451449924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,16384,0.7041858037312826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,16384,0.1916639937294854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,16384,0.6042790942721897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,51200,2.086512883504232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,65536,2.4135803646511502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,12288,0.14696444405449763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,12288,0.5291084183586968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,10240,0.3367244402567546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,10240,0.12753332985772028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,8192,0.3165244526333279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,12288,0.5735351244608561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,51200,2.3744818369547525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,8192,0.10272977749506633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,7168,0.275927119784885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,7168,0.09259288840823704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,8192,0.39710044860839844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,6144,0.21626667181650797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,6144,0.08771733442942302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,7168,0.28454844156901044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,5120,0.2021404504776001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,6144,0.2275786664750841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,5120,0.07063288821114434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,65536,3.514442655775282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,4096,0.1583368910683526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,3584,0.11534933249155681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,5120,0.19059377246432832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,4096,0.06551822026570638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,3072,0.1124746667014228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,3584,0.05990844302707248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,4096,0.15006044175889757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,2560,0.10243822468651666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,3584,0.13409066200256348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,2048,0.0718782212999132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,2560,0.044347554445266724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,1536,0.05755911270777384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,3072,0.1142444478140937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,2560,0.10196355316374038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,2048,0.035655998521380954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,1024,0.04494577646255493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,768,0.03012977706061469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,2048,0.08858933051427205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,10240,0.38680799802144367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,1024,0.024039111203617517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,512,0.026479111777411565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,768,0.022447999980714586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,1024,0.049901333120134145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,256,0.01798488861984677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,512,0.019918221566412184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,768,0.03965422179963853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,128,0.013882666826248169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,512,0.032615999380747475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,256,0.01833599971400367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,64,0.012083555261294046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,12288,32,0.0122079998254776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,128,0.018214222457673814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,128,0.0277511113219791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,64,0.028064890040291682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,32,0.027493332823117573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,3072,0.047037333250045776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,12288,1536,0.029260443316565618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,1536,0.06714933448367648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,65536,0.7267920176188151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,12288,256,0.028855111863878038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,16384,0.4754568735758464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,51200,0.5599102444118923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,16384,0.16929511229197183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,51200,1.4014435874091253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,12288,0.38401420911153156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,12288,0.13006756040785047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,65536,2.095822228325738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,10240,0.2938782109154595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,12288,0.41559555795457626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,10240,0.10991377962960137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,8192,0.2688719961378309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,10240,0.4427902168697781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,8192,0.08645333184136285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,7168,0.23538133833143446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,8192,0.2660408814748128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,51200,2.262280993991428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,7168,0.0785706639289856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,65536,2.658435609605577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,5120,0.16391644212934706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,6144,0.0701342225074768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,6144,0.19492177168528238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,5120,0.06176622046364678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,4096,0.1364391114976671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,5120,0.15604266855451795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,4096,0.049736890527937144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,4096,0.13375377655029297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,3072,0.09238221910264756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,16384,0.7370150883992513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,3584,0.11157245106167263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,2560,0.07573333051469591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,3072,0.03906933466593424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,3072,0.09630844328138564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,2048,0.060475554731157094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,6144,0.19108800093332926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,2560,0.08415466547012329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,1536,0.0476399991247389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,7168,0.2876017888387044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,2048,0.03066133459409078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,2048,0.06877244181103177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,1024,0.03452533483505249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,1536,0.024586665961477492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,1536,0.05528088741832309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,768,0.031750222047170006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,1024,0.021166221963034734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,512,0.026410667432679072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,1024,0.04186311033036974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,768,0.01906755566596985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,3584,0.04422488808631897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,256,0.020247111717859905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,512,0.017479111750920612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,128,0.01240711079703437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,512,0.027948445743984644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,256,0.016122667325867545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,64,0.010701333483060202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,256,0.024699555502997503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,128,0.015855999456511605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,10240,2560,0.035353776481416486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,128,0.023735110958417256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,64,0.0240346673462126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,32,0.02369600037733714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,3584,0.1262133386400011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,10240,768,0.03480089041921828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,65536,0.6019911236233181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,10240,32,0.011115555961926779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,16384,0.3897253407372369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,51200,0.48608891169230145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,51200,1.3926898108588324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,16384,0.14090222782558864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,12288,0.2855386734008789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,65536,1.637229283650716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,16384,0.5101786719428169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,12288,0.13067377938164607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,10240,0.2779937850104438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,12288,0.3378675513797336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,10240,0.10676711135440403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,8192,0.21557154920366076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,51200,1.6890764236450195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,8192,0.07351377937528822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,7168,0.1912471055984497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,65536,2.0347254011366105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,10240,0.3304639922247993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,7168,0.07089066505432129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,8192,0.23702756563822427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,5120,0.13239199585384792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,6144,0.05877421961890327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,7168,0.23645512262980142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,5120,0.05125244458516439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,6144,0.15841244326697454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,3584,0.09363022115495469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,4096,0.0410497784614563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,3072,0.07835288842519124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,4096,0.11088089148203532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,3584,0.03721333212322659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,3584,0.0941022237141927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,3072,0.033666667011049055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,3072,0.08000710937711927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,2048,0.055517332421408765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,2560,0.03045511245727539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,1536,0.04071377714474996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,2048,0.02571733295917511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,2048,0.057038221094343394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,1024,0.027679110566775005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,1536,0.02152266601721446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,1536,0.04603999853134155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,6144,0.16983289188808867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,768,0.02074666652414534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,1024,0.017948443690935772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,512,0.01643022232585483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,768,0.01647644407219357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,4096,0.11103288994895087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,256,0.012490666574902005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,768,0.028904000918070476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,5120,0.13298043939802381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,512,0.015156444576051501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,128,0.009956444303194681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,512,0.023407111565272015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,256,0.014087110757827759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,64,0.008960888617568547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,256,0.020603555772039626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,32,0.009004444711738164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,8192,2560,0.06681244240866767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,128,0.01995022263791826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,64,0.019938665959570143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,2560,0.06970133384068807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,32,0.01994311147265964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,8192,1024,0.035029331843058266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,65536,0.5056799782647027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,51200,1.001974211798774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,8192,128,0.013983110586802164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,16384,0.37598488065931535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,65536,1.5086409250895183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,51200,0.4190559917026096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,16384,0.12314132849375407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,12288,0.285720878177219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,12288,0.11258844534556071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,10240,0.22437689039442274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,65536,1.7675804562038844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,51200,1.5267084969414606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,16384,0.3871075577206082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,8192,0.17602845033009848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,10240,0.0962631106376648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,12288,0.3701351218753391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,7168,0.15469777584075928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,8192,0.06506044334835477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,6144,0.14308088355594212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,7168,0.06629688872231378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,5120,0.11069155401653713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,8192,0.21942667166392008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,6144,0.05667733483844333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,7168,0.15801955593956843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,4096,0.07927199867036608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,6144,0.1457386679119534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,5120,0.04660711023542616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,4096,0.041378666957219444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,3072,0.06979466809166802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,5120,0.12824800279405382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,3584,0.03362044360902574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,2560,0.05403288867738512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,3584,0.08218577835294935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,3072,0.031050665511025324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,2048,0.041542222102483116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,3072,0.07331822315851848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,2560,0.027719999353090923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,1536,0.033519999848471746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,2560,0.061843554178873696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,1024,0.02329066726896498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,2048,0.02344888945420583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,2048,0.05222488774193657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,1536,0.020278222031063504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,768,0.01830933325820499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,1536,0.0421688887808058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,1024,0.016893333858913846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,512,0.014653333359294467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,768,0.015630222029156156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,1024,0.031911111540264554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,256,0.010659555594126383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,10240,0.2231848769717746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,512,0.014194667339324951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,128,0.009311111436949836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,512,0.021176889538764954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,64,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,3584,0.08272888925340441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,128,0.01329066687160068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,7168,32,0.008048888709810045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,128,0.018061333232455783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,4096,0.09499822060267131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,64,0.018220444520314533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,32,0.018138666947682697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,65536,0.4794613520304362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,768,0.026501332720120747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,51200,0.7995208634270562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,65536,1.0842168596055772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,7168,256,0.013236444857385425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,7168,256,0.018562666243977018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,16384,0.3072444332970513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,16384,0.11841777960459392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,12288,0.2318951023949517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,51200,0.38122222158643937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,10240,0.18734578291575113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,12288,0.0850284430715773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,51200,1.1988159815470378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,16384,0.3386382261912028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,65536,1.553000873989529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,10240,0.08147644334369235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,7168,0.12027555041842991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,8192,0.14952622519599065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,10240,0.19200532966189912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,8192,0.05844089057710436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,7168,0.05349422163433499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,6144,0.10414577855004205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,5120,0.09780533446205987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,8192,0.16035289234585232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,6144,0.04580977890226576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,7168,0.15581156147850886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,4096,0.06837955448362562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,5120,0.04010222355524699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,3584,0.06665955649481879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,6144,0.13391110632154676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,4096,0.032927112446890935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,3072,0.05503999855783251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,5120,0.1021804412206014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,3584,0.030977778964572485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,4096,0.08231288856930204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,2560,0.0451822214656406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,3584,0.0719955563545227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,2048,0.03680533170700073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,3072,0.06395110819074842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,2560,0.024725332856178284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,1536,0.028016000986099243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,2560,0.054711110062069364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,1024,0.020019556085268658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,1536,0.01800355480776893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,2048,0.045383999745051064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,768,0.01590222285853492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,1536,0.036329779360029436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,12288,0.23207555876837838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,512,0.013175110850069257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,1024,0.015395555231306287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,1024,0.02810133496920268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,256,0.009685333404276106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,768,0.014372444815105863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,768,0.02347822156217363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,128,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,256,0.012262221839692859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,256,0.016827555166350473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,64,0.006870222174459034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,128,0.012376888758606382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,6144,32,0.0070053330726093715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,128,0.016195555528004963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,3072,0.02683288852373759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,64,0.01631733278433482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,32,0.016194666425387066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,2048,0.021570665968788996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,65536,0.4271715482076009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,6144,512,0.013402666482660504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,6144,512,0.018925334016482036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,65536,1.025411605834961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,51200,0.9262320200602213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,16384,0.2857288784450955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,51200,0.3379795551300049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,16384,0.10405777560340033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,12288,0.19694222344292533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,51200,1.0014151467217338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,16384,0.26593245400322807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,65536,1.398514641655816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,10240,0.17512267165713838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,8192,0.1339199940363566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,10240,0.06369600031110975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,12288,0.2426373428768582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,7168,0.12251733409033881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,8192,0.056709335909949414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,10240,0.1713333394792345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,6144,0.09629244274563259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,8192,0.1347502205106947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,5120,0.07411199808120728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,6144,0.04593422346644931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,7168,0.1348008844587538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,5120,0.035248001416524254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,6144,0.12123821841345893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,3584,0.05897421969307793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,5120,0.09049333466423883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,4096,0.030214223596784804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,3072,0.048471109734641186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,3584,0.02720355490843455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,2560,0.0439280006620619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,3072,0.024816888901922438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,3072,0.05564888980653551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,2048,0.03392533461252848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,2560,0.022138666775491502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,12288,0.07640088929070367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,2560,0.04692622356944614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,1536,0.02681866619322035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,2048,0.01926399932967292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,1024,0.018608000543382432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,1536,0.01680177781316969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,1536,0.03184355629814996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,7168,0.04592533244027031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,1024,0.014365333649847241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,768,0.015220445063379077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,512,0.011167111496130625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,1024,0.024473778075642053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,768,0.013305777476893531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,256,0.0114702218108707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,768,0.021176000436147053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,512,0.012210666305489011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,128,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,512,0.016579555140601266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,4096,0.0710115565194024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,256,0.011585777832402123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,64,0.006464889066086874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,5120,128,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,256,0.015127110812399121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,3584,0.0638151102595859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,32,0.00674044465025266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,128,0.014485332700941296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,64,0.014558222558763293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,32,0.014434667097197639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,5120,2048,0.04009955459170871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,65536,0.3577093283335368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,16384,0.21566222773657906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,65536,0.749482684665256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,5120,4096,0.05922933419545492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,51200,0.7014435662163628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,12288,0.17154399553934732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,16384,0.08304444286558363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,16384,0.2397680017683241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,10240,0.13271110587649876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,51200,0.803218682607015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,12288,0.07445866531795926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,12288,0.22554487652248809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,8192,0.11166221565670437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,10240,0.05493155452940199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,7168,0.08435289065043132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,8192,0.049399111005995006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,6144,0.08126755555470784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,8192,0.1162231100930108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,7168,0.04102133380042182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,7168,0.09911733203464085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,5120,0.06581066714392768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,6144,0.035736001200146146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,4096,0.05614844295713636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,6144,0.08701333072450425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,51200,0.28325067626105416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,3584,0.047889779011408486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,5120,0.0731368859608968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,4096,0.02600355611907111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,4096,0.058152887556287974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,3072,0.039376888010236956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,3584,0.023585778143670823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,2560,0.03252977795071072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,65536,1.0814150704277885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,3584,0.051979555024041064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,3072,0.021534222695562575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,2048,0.026170666019121807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,1536,0.021112889051437378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,3072,0.04612444506751167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,2560,0.019261333677503798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,2048,0.016925333274735343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,1024,0.015972443752818637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,10240,0.1409626669353909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,2560,0.040001776483323835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,768,0.013283555706342062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,2048,0.03381866547796462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,512,0.010024888647927178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,1024,0.012513777448071374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,1536,0.026879999372694228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,768,0.012042666474978128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,1024,0.02102222210831112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,512,0.011350221931934357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,128,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,512,0.014465777410401238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,5120,0.0306275553173489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,256,0.013076444466908773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,64,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,128,0.010576888918876648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,128,0.01278222186697854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,64,0.012395555774370829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,32,0.012454222473833295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,1536,0.01442399952146742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,65536,0.2932159900665283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,256,0.007585778004593319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,4096,768,0.018160889546076458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,4096,256,0.010388444695207808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,16384,0.14488977856106228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,51200,0.46381155649820965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,51200,0.23701511489020455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,65536,0.9383403990003797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,12288,0.13291911284128824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,16384,0.07164355781343248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,10240,0.11565244197845459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,16384,0.1979128917058309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,12288,0.06117688947253757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,8192,0.09716533290015326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,12288,0.15004800425635442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,51200,0.7321893374125162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,10240,0.05125688844256931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,7168,0.08265866835912068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,10240,0.1252817842695448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,8192,0.038132445679770574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,6144,0.06848444541295369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,8192,0.10792799790700276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,7168,0.03728888763321771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,5120,0.0573600000805325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,6144,0.030471109681659277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,4096,0.044659554958343506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,7168,0.09613955683178371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,5120,0.027832001447677612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,6144,0.07641066445244683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,5120,0.06573155853483412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,3072,0.03341333402527703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,3584,0.021010667085647583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,4096,0.05253155695067512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,2560,0.027718222803539697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,4096,32,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,3072,0.019184889064894784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,2048,0.0232595553000768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,3072,0.042034665743509926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,2560,0.01719999975628323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,1536,0.018548443913459778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,2560,0.036257776949140764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,2048,0.015402666396564908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,1024,0.013571555415789286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,2048,0.031195554468366835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,1536,0.024697777297761705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,1024,0.011953777737087674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,1024,0.01958666741847992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,768,0.011040888726711273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,3584,0.037498666180504694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,512,0.008623110751310984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,768,0.011114666859308878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,4096,0.02312444481584761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,256,0.006566222343179915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,768,0.016749334004190233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,512,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,3584,0.04817422231038412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,128,0.005756444401211209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,512,0.013455111119482251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,256,0.009740444521109263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,64,0.005195555587609609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,256,0.012044444680213928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,32,0.005532444351249271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,128,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,128,0.011682666838169098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,64,0.011699555648697747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3584,32,0.011688888900809817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3584,1536,0.013790222505728403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,65536,0.487018691168891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,65536,0.27872443199157715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,51200,0.3790880044301351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3584,65536,0.7107653088039823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,16384,0.14287732707129583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,51200,0.22021599610646567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,12288,0.1126355595058865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,16384,0.06844177511003283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,10240,0.09846933020485772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,12288,0.053215109639697604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,65536,0.7846719953748914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,16384,0.1741511159472995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,8192,0.07879910866419475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,51200,0.6270497639973959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,12288,0.14035644796159533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,10240,0.047713776429494224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,7168,0.06714489062627156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,8192,0.03639288743336996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,6144,0.056348443031311035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,10240,0.11387999852498372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,7168,0.033775111039479576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,5120,0.04755377769470215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,8192,0.09001600080066258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,4096,0.03974044322967529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,6144,0.029339555237028334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,7168,0.08407288789749146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,3584,0.03300266795688205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,5120,0.026208889153268602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,5120,0.059454222520192467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,3072,0.029829333225886028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,4096,0.04793511165512932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,3584,0.020263110597928364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,2560,0.024696000748210486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,3072,0.018412444326612685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,3584,0.043502221504847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,2048,0.02001333402262794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,1536,0.016732444365819294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,3072,0.037950221035215594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,2560,0.016713778177897137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,2048,0.01497244503762987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,1024,0.012153777811262341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,2560,0.03254755669169956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,768,0.010477333433098264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,1536,0.013456888496875763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,1024,0.011704000333944956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,1536,0.022266666094462078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,512,0.008970666262838576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,1024,0.017487999465730455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,768,0.015171556009186639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,512,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,512,0.012433778080675336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,128,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,256,0.009681777821646797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,256,0.011142222417725457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,6144,0.06899288627836439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,4096,0.022096888886557683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,128,0.010728889041476779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,32,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,64,0.010709332923094431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,32,0.010737777584128909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,3072,2048,0.02770488957564036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,65536,0.5187502437167698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,768,0.010975999964608086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,65536,0.2534088823530409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,256,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,51200,0.3504657745361328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,16384,0.12918133205837673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,51200,0.20109689235687256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,65536,0.6855671140882703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,12288,0.09600622124142116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,16384,0.06067022350099352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,3072,64,0.004577777865860197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,16384,0.15296444627973768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,3072,128,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,12288,0.04722133278846741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,8192,0.07747111055586073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,10240,0.04332177837689718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,51200,0.591426690419515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,12288,0.1144533289803399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,7168,0.06616533464855619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,8192,0.033909334076775446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,10240,0.09672444396548802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,6144,0.053704000181621976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,8192,0.07744622230529785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,5120,0.04610844453175863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,6144,0.027349332968393963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,7168,0.07331111033757527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,4096,0.03584622343381246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,5120,0.023390221926901076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,3584,0.031826668315463595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,5120,0.051022221644719444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,4096,0.020568889048364427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,3072,0.027642667293548584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,4096,0.04229866796069675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,3584,0.01835466590192583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,3584,0.037440889411502414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,2560,0.022827555735905964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,3072,0.016996444927321542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,2048,0.020217777954207528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,3072,0.03367644548416138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,2560,0.014997333288192749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,10240,0.09343200259738499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,1536,0.015645333462291293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,2048,0.014728888869285583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,1024,0.012102222277058495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,2048,0.024665777881940205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,1536,0.012064889073371887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,768,0.00960799968904919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,1536,0.019923556182119582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,7168,0.03024266825781928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,1024,0.010674667027261523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,1024,0.015618667006492615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,512,0.007359999749395583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,768,0.010047111246320937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,6144,0.06015288829803467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,512,0.009534222384293875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,128,0.004842666702138053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,512,0.011176000038782755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,256,0.009304000271691216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,256,0.01036711119943195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,64,0.004534222185611725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2560,128,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,32,0.004596444467703502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,128,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,64,0.009687999884287516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,32,0.009716444545321995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,2560,0.02868177824550205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,65536,0.43005866474575466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,65536,0.20938577916887072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,51200,0.3358302116394043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,16384,0.1008533305592007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,51200,0.19117421574062773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,65536,0.5268097983466254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,12288,0.07168177763621013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2560,256,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,16384,0.06322044134140015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,10240,0.0590462221039666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,16384,0.13027732902103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,51200,0.4083111021253798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,8192,0.05099022057321337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,7168,0.04154577851295471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,12288,0.10495911041895549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,10240,0.0823884473906623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,8192,0.029658665259679157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,6144,0.03589866558710734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,8192,0.0680053300327725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,7168,0.027748444014125403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,5120,0.03312622176276313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,7168,0.05995644463433159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,6144,0.024142222272025213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,4096,0.02525955604182349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,6144,0.05163644419776069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,5120,0.021320000290870667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,5120,0.04498666524887085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,4096,0.01817599932352702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,3072,0.019860444797409903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,4096,0.03636088967323303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,3584,0.016859556237856548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,3584,0.03233333428700765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,3072,0.015545777148670621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,3072,0.029013332393434312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,2048,0.014927999840842353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2560,768,0.013779555757840475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,12288,0.042633778519100614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,2560,0.024991111622916326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,10240,0.04279999931653341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,2048,0.013067555924256643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,1024,0.00924444446961085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,2048,0.021185777253574792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,1536,0.011408000356621213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,768,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,1024,0.01036888857682546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,1024,0.013811555173661975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,512,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,768,0.009682666924264696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,3584,0.023112888137499493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,768,0.012111110819710625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,256,0.004986666556861666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,512,0.009226666556464301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,512,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,128,0.0041884444653987885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,256,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,256,0.009249777429633671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,64,0.0038488888078265717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,128,0.008614222208658854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,2048,2560,0.014463999205165438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,128,0.008747555315494537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,32,0.004222222086456087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,1536,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,64,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,32,0.008712000317043727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,65536,0.27859022882249623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,2048,1536,0.017509332961506315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,65536,0.16363378365834555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,51200,0.2527528868781196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,16384,0.07945866717232598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,12288,0.058973332246144615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,51200,0.13305066691504583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,16384,0.04622933268547058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,10240,0.0539644459883372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,51200,0.3338622252146403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,12288,0.03565333286921183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,8192,0.04059999850061204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,12288,0.08047911193635729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,2048,2560,0.017431111799346078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,10240,0.03119200136926439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,7168,0.03491111265288459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,10240,0.06890310843785603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,8192,0.024307555622524683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,6144,0.03143289022975498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,8192,0.055890666113959424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,7168,0.02218311031659444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,5120,0.02607111136118571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,7168,0.049674666590160794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,6144,0.019884443945354886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,4096,0.021544888615608215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,6144,0.04351733459366692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,5120,0.01791288952032725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,3584,0.01997066703107622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,5120,0.036864889992607966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,4096,0.01571911076704661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,3072,0.01755199995305803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,3584,0.014478221535682678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,2560,0.015484443969196744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,3072,0.013487999637921652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,65536,0.4238995446099176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,2048,0.012687111066447364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,3072,0.024474667178259954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,2560,0.012256888879670037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,16384,0.1081004473898146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,1536,0.010426667001512317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,1024,0.008258666429254744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,2048,0.01109155515829722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,2560,0.021095110310448542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,2048,0.0181057784292433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,1536,0.010089777410030365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,1024,0.009377777576446533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,1536,0.01516355574131012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,512,0.005867555737495422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,1024,0.012496888637542725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,768,0.008996444443861643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,256,0.004548444516128964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,768,0.010608889162540436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,512,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,128,0.0038293335172865125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,256,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,512,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,64,0.0035386664999855887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1536,128,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,32,0.0038328886859946777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,128,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,64,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,4096,0.03081511126624213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,65536,0.19987466600206164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,32,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,3584,0.027627555860413447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,51200,0.18146222167544893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,16384,0.054577777783075966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,65536,0.15772533416748047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,51200,0.12573688560061985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,12288,0.042167110575569995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,65536,0.3244853284623888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,16384,0.04225155711174011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1536,768,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,10240,0.036892443895339966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,51200,0.25602221488952637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,8192,0.030360000001059637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,12288,0.03304622239536709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,16384,0.08428266975614761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,7168,0.025672889418072168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,10240,0.029263112280103896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,12288,0.0639831092622545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,8192,0.02249777813752492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,6144,0.022448889083332483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,10240,0.05367911193105909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,8192,0.044234666559431285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,7168,0.0209840006298489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,6144,0.018346667289733887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1536,256,0.008105778031879002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,4096,0.015409777561823526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,5120,0.016701334052615695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,6144,0.034653332498338484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,3584,0.013949332965744866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,4096,0.01459377838505639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,3072,0.012056888805495368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,4096,0.024624889095624287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,3584,0.013789333403110504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,3584,0.022115555074479844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,3072,0.012740444805887012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,3072,0.020049777295854356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,2048,0.00905955582857132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,2560,0.011592888997660743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,2560,0.017441777719391715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,1536,0.0075235557225015424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,2048,0.010537777509954242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,2048,0.015142222245534262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,1024,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,1536,0.009671111073758867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,5120,0.019310222731696237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,768,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,7168,0.03936800029542711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,1024,0.008985777695973715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,512,0.004298666699065102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,1024,0.010325333310498131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,768,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,5120,0.029639999071757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,256,0.003522666792074839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,512,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,128,0.0033235556135574975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,2560,0.010864888628323873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,512,0.007663999994595845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,256,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,64,0.002830222249031067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,256,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,1024,32,0.0031475555151700974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,1024,128,0.007631111476156447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,128,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,65536,0.15533422099219427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,64,0.006754666566848755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,32,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,51200,0.13071910540262857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,65536,0.1556560066011217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,16384,0.04702044526735941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,1536,0.01272266689274046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,51200,0.12343377537197536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,12288,0.032348444064458214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,10240,0.025618667403856914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,16384,0.04055022199948629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,51200,0.21978133254581025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,8192,0.022870222727457683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,12288,0.03247022297647264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,16384,0.073308441374037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,12288,0.055480887492497764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,1024,768,0.009006222089131674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,8192,0.022487999664412603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,10240,0.047167999876870044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,6144,0.017436444759368896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,7168,0.020395555429988436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,8192,0.038959999879201256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,5120,0.014366222752465142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,7168,0.03502310978041755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,6144,0.018365333477656048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,4096,0.012416889270146688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,6144,0.030821333328882854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,5120,0.016556445095274184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,3584,0.011218667030334473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,5120,0.026231999198595684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,4096,0.014570666684044732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,3072,0.00997155573632982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,3584,0.013456888496875763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,4096,0.021721776988771226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,2560,0.009024000002278222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,3584,0.019669334093729656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,2048,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,3072,0.01805688937505086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,2560,0.011699555648697747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,1536,0.00656444455186526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,2560,0.015588444140222339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,2048,0.010497777826256223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,1024,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,2048,0.01370044466522005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,1536,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,768,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,7168,0.01981599960062239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,1536,0.011586666935020022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,1024,0.009019555317031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,768,0.00851733320289188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,768,0.008376888930797577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,256,0.003168888803985384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,512,0.008016889293988546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,512,0.007249777515729268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,128,0.002918222298224767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,256,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,256,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,64,0.0028355556229750314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,128,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,3072,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,32,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,65536,0.2782675425211589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,64,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,65536,0.13513422012329102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,32,0.0063484443558586975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,10240,0.030247999562157526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,51200,0.11072088612450494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,65536,0.11174755626254612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,16384,0.035912000470691256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,768,1024,0.009745777481132084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,65536,0.2359466552734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,51200,0.09049510955810547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,12288,0.025403555896547105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,16384,0.03030666708946228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,51200,0.18592000007629395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,10240,0.021908443835046556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,16384,0.06132444408204821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,12288,0.024209777514139812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,8192,0.018352000249756705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,10240,0.021852445271280076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,7168,0.015992888145976596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,8192,0.016751999656359356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,8192,0.03339555528428819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,6144,0.01425066590309143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,7168,0.015542222393883599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,7168,0.029882666137483384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,5120,0.012416000167528788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,6144,0.014142221874660917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,6144,0.026335999369621277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,4096,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,5120,0.012840000291665396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,768,512,0.0038355555799272326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,3584,0.00867911097076204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,4096,0.011491555306646558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,4096,0.01928622192806668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,3072,0.008070222205585903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,3584,0.011055111057228513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,3584,0.017433777451515198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,2560,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,3072,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,12288,0.04732622371779548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,2048,0.006148444281684027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,3072,0.01567555632856157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,2560,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,768,128,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,10240,0.040179553959104754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,1536,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,2560,0.013786666923099093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,1024,0.0041982221106688184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,2048,0.009336888790130615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,768,0.0035439998739295532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,1536,0.009031111167536842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,1024,0.008069333102968004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,2048,0.012158222496509552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,512,0.0032213332338465583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,1536,0.01034755590889189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,256,0.0028488888508743713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,1024,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,128,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,768,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,256,0.00702400008837382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,512,0.006859555426571105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,64,0.002415111081467734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,512,32,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,128,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,5120,0.022774222824308608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,128,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,65536,0.07584444681803386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,51200,0.05946844153934055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,32,0.005695111221737332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,65536,0.11139021979437934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,16384,0.02048711147573259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,65536,0.18950488832261828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,51200,0.08983200126224095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,51200,0.1488773292965359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,12288,0.016960889101028442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,16384,0.02975733412636651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,10240,0.015509333875444202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,12288,0.02327466673321194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,12288,0.03880000114440918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,768,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,8192,0.01313866674900055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,10240,0.021376889612939622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,512,512,0.007328000333574083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,10240,0.03342577815055847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,8192,0.01661422186427646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,256,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,6144,0.010119111173682744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,7168,0.015436444017622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,8192,0.027526222997241553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,5120,0.008753777378135258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,7168,0.024838222397698298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,6144,0.014037332601017423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,512,64,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,5120,0.01277511070171992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,6144,0.022042666872342426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,3584,0.006827555596828461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,5120,0.01905777719285753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,4096,0.011436444189813403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,4096,0.016376000311639573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,3584,0.015049777097172208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,16384,0.05045333173539904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,2560,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,3072,0.010425777898894416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,2560,0.009872000250551436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,2048,0.004639111045334074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,7168,0.011536888778209686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,2560,0.012220444778601328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,1536,0.004192000048028098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,2048,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,2048,0.01096088853147295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,1024,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,1536,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,4096,0.007241778075695038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,1536,0.009304000271691216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,768,0.0031724443866146933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,1024,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,3072,0.006211555666393704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,512,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,1024,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,3584,0.011071110765139261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,256,0.002510222295920054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,768,0.007242666350470648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,128,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,256,0.007003555695215861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,3072,0.013757333159446716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,64,0.0021048889805873237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,128,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,256,32,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,128,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,65536,0.05853688716888428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,64,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,32,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,51200,0.04714666803677877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,65536,0.11099111371570164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,65536,0.18855110804239908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,16384,0.016176889340082806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,51200,0.08973955445819431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,12288,0.012811555630630918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,51200,0.14840622742970785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,16384,0.05022222134802076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,10240,0.011401777466138205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,12288,0.02324177821477254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,512,0.007344000041484833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,12288,0.038753777742385864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,8192,0.011904888682895236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,512,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,10240,0.021498666869269475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,7168,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,256,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,10240,0.032736887534459434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,8192,0.016686222619480558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,6144,0.009377777576446533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,5120,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,7168,0.015425778097576566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,8192,0.027061333258946735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,6144,0.013909333282046847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,4096,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,7168,0.024373332659403484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,5120,0.01295733369059033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,6144,0.021640888518757288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,5120,0.01889866590499878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,4096,0.011422221859296164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,3072,0.005635555419656966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,3584,0.010951110886202918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,4096,0.016211556063758004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,2560,0.005055111315515307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,3584,0.014820444915029736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,3072,0.01038933296998342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,2048,0.00452533322903845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,16384,0.02899644441074795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,3072,0.013651555611027611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,2560,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,1536,0.003847111016511917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,256,768,0.007848000360859765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,1024,0.003160888950030009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,2048,0.009317333499590555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,768,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,2048,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,1024,0.00816888858874639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,1536,0.009135111338562435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,512,0.002537777854336633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,768,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,1024,0.007954667011896769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,256,0.002390222210023138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,768,0.006990222467316522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,512,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,128,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,512,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,128,0.007170666423108842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,256,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,32,0.00221422231859631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,128,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,128,3584,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,65536,0.05397422115008036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,51200,0.04275911053021749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,64,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,16384,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,65536,0.1893448829650879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,51200,0.14818755785624185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,12288,0.012183999849690331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,16384,0.050027555889553495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,10240,0.010944888823562197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,2560,0.012101333174440594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,8192,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,7168,0.00850044439236323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,1536,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,6144,0.00793599999613232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,12288,0.0380257765452067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,5120,0.008632000121805403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,10240,0.03218577636612786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,8192,0.026590221458011206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,4096,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,7168,0.024114666713608637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,6144,0.02147644427087572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2305,128,256,0.0070017774899800616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,3072,0.005882666756709416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,5120,0.018927110566033255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,4096,0.016169778174824186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,3584,0.014813333749771118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,3072,0.013554666605260638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,1536,0.0038577777643998465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,128,32,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,2560,0.012048000262843238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,1024,0.0032071111102898917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,768,0.0028604444944196274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,1536,0.009390222529570261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,1024,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,512,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,256,0.0021582223061058256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,768,0.007039111521508958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,128,0.001976888833774461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,512,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,256,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,64,0.00180177783800496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,32,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,128,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,65536,0.048151999711990356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,2560,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,64,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,51200,0.03896888759401109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,2048,0.004491555607981152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,16384,0.01036533299419615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,12288,0.008723555339707268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,65536,0.18963555494944254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,2048,0.010727110836240979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,10240,0.008178667061858708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,16384,0.04997333221965366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,8192,0.007496000164084964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,7168,0.0070062221752272705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,12288,0.037795556916130915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,6144,0.009959999885823991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,5120,0.0173688895172543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,10240,0.03251111176278856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,7168,0.0240320000383589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,6144,0.02162933349609375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,3584,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,5120,0.018862222631772358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,3072,0.006715555571847492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,4096,0.016192888220151264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,2560,0.009712888962692684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,3072,0.013502221968438892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,2048,0.004398222184843487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,2560,0.012059555285506778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,1536,0.003834666477309333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,51200,0.14839288923475477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,2048,0.01074399964676963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,1024,0.003181333343187968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,768,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,1536,0.008985777695973715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,512,0.002471999989615546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,8192,0.02667733364635044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,1024,0.00776622196038564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,4096,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,256,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,768,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,128,0.0018560000591807896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,64,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,32,32,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,512,0.00601244428091579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2305,64,3584,0.006260444306664997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,256,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,128,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,64,0.005294222384691238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,32,0.004961777892377642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,64,32,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2305,32,3584,0.014831110835075378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,16384,0.9200479719373914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,16384,2.8685351477728953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,16384,3.266738679673937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,51200,3.444053437974718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,16384,2.761403613620334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,12288,2.2428595225016275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,12288,2.260808096991645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,12288,0.6707955466376411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,10240,1.9679465823703344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,10240,0.5757662455240885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,12288,2.003236346774631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,10240,1.8582827250162761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,10240,1.7028382619222004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,8192,1.625775973002116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,8192,1.6174329121907551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,51200,9.041378445095486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,8192,0.4497084617614746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,8192,1.369418674045139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,7168,0.4035324520534939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,7168,1.4096612930297852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,7168,1.2750604417588975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,6144,1.103260464138455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,6144,0.3564399878184001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,6144,1.0820862452189128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,7168,1.1419270833333333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,5120,1.03840086195204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,5120,0.9188612831963433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,5120,0.3125573264227973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,6144,1.009071985880534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,4096,0.7743217680189344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,4096,0.25512801276312935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,5120,0.7930115593804253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,3584,0.6653395758734809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,4096,0.7638337877061633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,3584,0.2317404482099745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,4096,0.6618515650431315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,51200,13.034555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,3584,0.7154746585422092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,3072,0.5480195681254069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,3072,0.200819558567471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,3584,0.5735093222724067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,2560,0.4748871061537001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,2560,0.1847848892211914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,2048,0.3716142177581787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,3072,0.5105857849121094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,2560,0.4622489081488715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,2560,0.44498491287231445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,1536,0.2753209008110894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,1536,0.29635821448432076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,3072,0.6311982472737631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,1536,0.1439137723710802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,1024,0.1685884396235148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,1536,0.2810640070173475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,51200,9.753836737738714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,1024,0.22473245196872285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,768,0.15494310855865479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,1024,0.24065423011779785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,768,0.08824444479412502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,512,0.11725510491265191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,768,0.1717217763264974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,2048,0.18046755260891387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,256,0.07846489217546251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,512,0.15846933258904353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,512,0.07601600223117404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,2048,0.35498489273919004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,512,0.14796800083584255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,2048,0.6117671330769857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,128,0.062000890572865806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,256,0.0680239995320638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,64,0.05787644121381971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,128,0.06526222493913439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,256,0.11024977763493855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,256,0.1365991168551975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,128,0.1298151148690118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,65536,32,0.057016889254252114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,65536,128,0.08329333199395074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,65536,1024,0.09668978055318196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,64,0.13103732797834608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,32,0.1318968931833903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,65536,768,0.2318577766418457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,51200,2.669527901543511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,65536,3.407772488064236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,16384,2.9379083845350475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,51200,7.238914489746094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,16384,2.6708123948838978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,16384,0.6981129116482205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,65536,9.00876702202691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,12288,1.905835469563802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,51200,10.221293131510416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,12288,0.5311208830939399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,12288,1.8929511176215277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,16384,2.1999564700656467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,10240,0.4495404561360677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,10240,1.5542177624172633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,51200,7.4894214206271705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,10240,1.5477191077338324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,8192,0.3651004367404514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,10240,1.3059679667154949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,12288,1.5797128677368164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,8192,1.4100284576416016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,8192,1.1509244706895616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,65536,13.60148705376519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,8192,1.0689475801255968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,7168,0.40343112415737575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,7168,1.2188071144951713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,6144,0.8435217539469401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,7168,1.0452328787909615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,65536,9.28683556450738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,6144,0.2839084466298421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,7168,0.8686969015333387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,6144,0.8833164638943143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,5120,0.263991117477417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,4096,0.5959902339511448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,5120,0.8945929209391276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,6144,0.759983115726047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,5120,0.8036470943027072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,4096,0.5946933428446451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,4096,0.21823467148674858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,5120,0.6434364318847656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,3584,0.595320012834337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,3584,0.18230666054619682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,3072,0.4556613498263889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,3584,0.5077555444505479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,4096,0.5096897549099392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,3584,0.4419608910878499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,3072,0.16198489401075575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,3072,0.44976531134711367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,2560,0.4161715507507324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,2560,0.138483550813463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,2048,0.2768790986802843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,3072,0.39815910657246906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,2560,0.36059199439154727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,1536,0.2074000040690104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,2048,0.11699644724527995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,2048,0.2984275552961561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,2560,0.3332551055484348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,1024,0.14172532823350695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,1536,0.09631644354926215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,1536,0.24120977189805773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,2048,0.395797332127889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,1024,0.07554488711886935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,768,0.11294933160146077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,1024,0.17715022299024794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,1024,0.1709146632088555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,512,0.08544800016615127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,1536,0.22365956836276582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,768,0.06784444385104708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,256,0.06251733170615302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,768,0.14189778433905706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,512,0.06167199876573351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,512,0.12160889307657878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,128,0.051238222254647144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,768,0.14514311154683432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,256,0.055460443099339805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,64,0.04625599914126926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,256,0.10812177922990586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,51200,32,0.047563556167814464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,51200,128,0.05266844564014011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,128,0.10286933183670044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,128,0.06742578082614474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,64,0.10395199722713894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,51200,32,0.10452266534169515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,512,0.1145004431406657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,51200,256,0.08820088704427083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,65536,0.9944933785332574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,51200,0.7895093493991427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,16384,0.8769662115308973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,51200,2.6008923848470054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,65536,2.937305874294705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,16384,0.2647751172383626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,51200,3.145018683539497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,12288,0.5644302368164062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,16384,0.8462888929578992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,12288,0.2065511147181193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,16384,0.6766782336764865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,10240,0.49673954645792645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,12288,0.584404468536377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,65536,4.170086330837674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,12288,0.48563554551866317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,10240,0.18460444609324136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,8192,0.3903644349839952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,8192,0.1389333274629381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,65536,3.0676070319281687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,8192,0.39064889483981663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,10240,0.4175573454962836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,51200,2.350842581854926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,7168,0.3727564546797011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,8192,0.3316391044192844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,6144,0.2857920063866509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,7168,0.3347457779778375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,6144,0.09603289100858901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,5120,0.23945599132113984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,7168,0.30482933256361217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,6144,0.3547946612040202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,6144,0.32489776611328125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,4096,0.1865004433525933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,5120,0.09753421942392985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,5120,0.23396889368693033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,3584,0.16715733210245767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,4096,0.19796978102789986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,5120,0.21526488992902967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,10240,0.4882462289598253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,4096,0.17869510915544298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,3584,0.16428800423940024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,3584,0.06309244367811415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,7168,0.11111556159125434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,2560,0.11272622479332818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,3584,0.1611608929104275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,3072,0.05718133184644911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,3072,0.1437564425998264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,3072,0.1406080060535007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,2048,0.09920977883868748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,2560,0.05054044392373827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,2560,0.1237671110365126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,2048,0.043772445784674756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,1536,0.06742133034600152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,2048,0.10183733039432102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,2048,0.10024799903233846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,1536,0.03577422102292379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,1024,0.05157866742875841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,4096,0.07608355416191949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,1536,0.08158933454089694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,768,0.036772443188561335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,1536,0.08398755391438802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,3072,0.13616000281439886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,1024,0.02995733420054118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,512,0.027676444914605882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,1024,0.06257688999176025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,1024,0.06392533249325223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,768,0.02751733362674713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,768,0.050035556157430015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,256,0.020074667202101815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,768,0.05504799882570902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,512,0.02517066730393304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,512,0.042862223254309766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,128,0.018236445056067575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,512,0.04522577921549479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,256,0.03821422325240241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,256,0.02312711046801673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,256,0.036129779285854764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,16384,128,0.022014222211307947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,32,0.01535377734237247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,128,0.03664711117744446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,128,0.027709333433045283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,64,0.03677333394686381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,16384,32,0.0372559991147783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,16384,64,0.015459555718633862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,16384,2560,0.11903999911414252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,65536,0.7619839774237739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,65536,2.0696640014648438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,51200,1.7850702073838975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,16384,0.7000257703993055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,51200,0.5982284545898438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,16384,0.2101271152496338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,12288,0.40383466084798175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,16384,0.6061573558383518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,16384,0.5238862037658691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,12288,0.4491582446628147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,65536,3.374729792277018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,12288,0.3903955618540446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,10240,0.361333343717787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,51200,2.525780359903971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,65536,2.4255803426106772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,10240,0.1184577809439765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,10240,0.3632444540659587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,8192,0.3342453373803033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,51200,1.8011343214246962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,10240,0.32621243264940053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,7168,0.23459556367662218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,7168,0.08713689115312363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,8192,0.27874755859375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,6144,0.1776684390174018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,12288,0.13822221755981445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,7168,0.31909333335028756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,7168,0.25757066408793133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,5120,0.16955910788642037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,6144,0.21535821755727133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,5120,0.18913155131869844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,6144,0.2012666728761461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,5120,0.06425422430038452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,4096,0.1450586716334025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,5120,0.16996355851491293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,3584,0.12312266561720107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,4096,0.054029332266913525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,8192,0.09659022092819214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,4096,0.1456177764468723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,8192,0.3165004518296984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,3584,0.04855555627081129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,4096,0.1388604508505927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,3584,0.12875999344719782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,2560,0.09011289146211411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,3584,0.13094488779703775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,3072,0.04370222157902188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,3072,0.11144622166951497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,2048,0.06755289104249743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,3072,0.11215378178490533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,2560,0.03990755478541056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,6144,0.07604089048173693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,2560,0.09609333011839126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,1536,0.05207555492719015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,2560,0.09423289034101699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,2048,0.03401333424780104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,1024,0.03756355577045017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,2048,0.07905599806043836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,1536,0.02754666739039951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,2048,0.08185866806242201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,768,0.027453333139419556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,1536,0.06333777639600965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,1536,0.06593866480721368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,1024,0.02271911170747545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,512,0.02150844368669722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,1024,0.048342221313052706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,768,0.02075199948416816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,1024,0.051510221428341336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,768,0.03906666570239597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,256,0.015512888630231222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,768,0.04446577694680956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,512,0.031706667608684964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,3072,0.10362755590014988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,512,0.03734044565094842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,256,0.01761866609255473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,256,0.028078221612506445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,64,0.01239999963177575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,256,0.030192000998391047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,32,0.012410666379663678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,128,0.0177208897140291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,128,0.027426666683620874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,12288,128,0.02331377731429206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,64,0.027036445008383855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,12288,32,0.02721333338154687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,12288,512,0.01868266695075565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,12288,128,0.014456000592973499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,65536,0.6406657960679796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,51200,1.6255822711520727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,65536,2.0729556613498263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,51200,0.5104862319098579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,16384,0.17381778028276232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,16384,0.5186115370856391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,65536,2.5457884470621743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,16384,0.4969813558790419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,12288,0.39084800084431964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,16384,0.43681245379977757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,51200,1.993485344780816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,12288,0.12053778436448838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,12288,0.3877822293175592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,10240,0.2851235601637098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,65536,1.97545772128635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,10240,0.10246044397354126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,12288,0.33676711718241376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,51200,1.548172420925564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,8192,0.23490844832526314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,10240,0.30465955204433864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,7168,0.19990932941436768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,10240,0.27571378813849556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,8192,0.2710862159729004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,7168,0.07413599888483684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,8192,0.23526310920715332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,7168,0.26585955089992946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,7168,0.21294844150543213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,5120,0.14537689420912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,6144,0.06512533293830024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,6144,0.1884240044487847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,4096,0.12090133296118842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,5120,0.05612711111704508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,5120,0.15899022420247397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,5120,0.14577689435746935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,4096,0.046625776423348315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,3584,0.10197955369949341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,4096,0.12320000595516628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,3072,0.08464799986945258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,4096,0.11911288897196452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,3584,0.04176444477505154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,3584,0.10837866862614949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,8192,0.08311466375986735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,3072,0.03726933399836222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,3584,0.10929244756698608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,2560,0.07129066520267062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,3072,0.09509244230058457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,2048,0.056863109270731606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,6144,0.15050221814049616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,3072,0.09466755390167236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,2560,0.033759110503726535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,1536,0.04535377687878079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,2560,0.08187911245557997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,2048,0.06763466861512926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,2048,0.02879555688963996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,1024,0.02869066596031189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,1536,0.023708444502618577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,1536,0.05447466505898369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,6144,0.17204622427622476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,768,0.022452443838119507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,1536,0.057314667436811656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,1024,0.02032888929049174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,1024,0.041606220934126116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,512,0.017542221479945712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,768,0.018128888474570382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,1024,0.04621244470278422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,768,0.033987555238935686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,256,0.013346666263209449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,768,0.03935022155443827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,512,0.01628888812330034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,128,0.010660444696744284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,512,0.02739288906256358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,512,0.03300711181428697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,256,0.015470221638679504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,256,0.024068444967269897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,256,0.025514667232831318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,10240,128,0.015551111764378019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,32,0.009983110758993361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,128,0.02331111166212294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,128,0.019991111424234178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,64,0.02335022224320306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,2560,0.08294488986333211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,10240,32,0.02336799932850732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,10240,2048,0.0694720016585456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,65536,0.5314213434855143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,10240,64,0.009640889035330879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,65536,1.392067591349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,51200,1.0220710966322157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,16384,0.3509199884202745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,51200,0.42050578859117294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,16384,0.42365421189202207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,51200,1.7043715582953558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,16384,0.1450915601518419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,12288,0.23026932610405815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,16384,0.3581644429100885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,12288,0.09845955504311456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,10240,0.2101937797334459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,65536,1.6110613081190321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,12288,0.2718417909410265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,8192,0.1886604494518704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,10240,0.08335110876295303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,10240,0.257114675309923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,10240,0.22438311576843262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,8192,0.06747555732727051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,51200,1.2374471028645833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,8192,0.2011706696616279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,6144,0.14184444480472141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,65536,2.1856613159179688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,8192,0.19744000169965956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,7168,0.17566133870018852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,7168,0.16237333085801867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,5120,0.10906044642130534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,6144,0.15056888262430826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,6144,0.14071022139655218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,12288,0.32041245036655003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,4096,0.0972364412413703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,5120,0.045828445090187915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,5120,0.12624266412523058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,3584,0.06916266679763794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,5120,0.12102577421400283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,4096,0.03829777903027005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,4096,0.10037333435482448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,3072,0.06926844517389934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,7168,0.16660711500379774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,3584,0.034509334299299456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,4096,0.10046488708919948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,3584,0.08934311072031657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,2560,0.06173244449827406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,3584,0.09018222490946452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,3072,0.03149244520399306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,2048,0.04573511083920797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,3072,0.07826044162114461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,2560,0.028008000718222723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,3072,0.08181155390209623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,1536,0.035465776920318604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,6144,0.05301955673429701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,2048,0.024128889044125874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,2048,0.05675377448399862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,1024,0.02368533280160692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,1536,0.020728000336223178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,1536,0.04526488979657491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,1536,0.049343999889161855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,1024,0.01721955504682329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,1024,0.03479111194610596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,1024,0.038935999075571694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,768,0.028669334120220605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,768,0.015604444675975375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,512,0.015102222561836243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,768,0.034118221865759954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,7168,0.071033775806427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,256,0.010758221977286868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,512,0.014588443769348992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,512,0.023019555542204116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,2560,0.06804000006781684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,128,0.009753777749008602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,2560,0.06973689132266574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,256,0.02021422154373593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,256,0.013348444468445249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,64,0.008383110993438298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,2048,0.05922400289111667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,256,0.021928888228204515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,32,0.008582221964995066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,8192,128,0.013462222284740873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,128,0.01959022217326694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,128,0.01737777723206414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,8192,768,0.018837332725524902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,32,0.0196106665664249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,65536,0.49946313434176975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,8192,512,0.028259555498758953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,51200,0.9867768817477756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,65536,1.4639689127604167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,8192,64,0.01961955593691932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,16384,0.341048002243042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,51200,0.39102310604519314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,51200,1.4478862550523546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,12288,0.26739467514885795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,16384,0.1189368830786811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,65536,2.0267626444498696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,16384,0.3756079938676622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,16384,0.3189226786295573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,12288,0.08596266640557183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,65536,1.3972959518432617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,10240,0.24121244748433432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,12288,0.26889777183532715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,10240,0.22042489051818848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,8192,0.19522755675845674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,10240,0.08471644586986965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,51200,1.066147592332628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,10240,0.20176355044047037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,7168,0.16880977153778076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,8192,0.17698044247097441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,6144,0.1475937763849894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,8192,0.164974225891961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,7168,0.05748799774381849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,7168,0.16854311360253227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,6144,0.04892622099982368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,7168,0.14811555544535318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,5120,0.11684000492095947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,6144,0.13338400257958308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,6144,0.12752800517612034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,4096,0.09420710802078247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,5120,0.0421653323703342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,5120,0.11628711223602295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,5120,0.11097777552074856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,3584,0.07285244597329034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,4096,0.03485066692034403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,12288,0.27162310812208385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,3072,0.07385155227449205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,4096,0.09328977929221259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,3584,0.03124444352255927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,8192,0.07530222336451213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,3584,0.0797591143184238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,2560,0.05466399921311272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,3072,0.028377778000301782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,3584,0.0827546649509006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,3072,0.07027200195524427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,2048,0.044405334525638156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,2560,0.025553777813911438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,3072,0.07403200202518039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,2560,0.06074666976928711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,2048,0.022221333450741235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,2560,0.06443822383880615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,2048,0.05053244365586174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,1024,0.021108445194032457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,2048,0.05412711037529839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,1536,0.018401778406567044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,1536,0.04070577687687344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,768,0.019371555911170114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,1536,0.04456177684995863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,1024,0.03159555461671617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,512,0.014903111590279473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,1024,0.0361422234111362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,768,0.014532445205582513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,768,0.025966222087542217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,256,0.011752000285519494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,768,0.031168000565634832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,512,0.01347733371787601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,512,0.020943111843532983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,512,0.025397333833906386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,256,0.012459555433856117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,256,0.018368888232443068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,256,0.021386666430367365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,64,0.007860444486141205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,128,0.012661333713266583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,1536,0.033688889609442815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,128,0.01757955551147461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,7168,128,0.01719288859102461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,64,0.017900443739361234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,32,0.017771555317772757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,7168,1024,0.015491555134455362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,7168,4096,0.08955555491977268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,65536,0.41540710131327313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,65536,1.0133146709865994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,51200,0.7720453474256727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,32,0.008276444342401292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,16384,0.2711004416147868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,51200,0.31600356101989746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,16384,0.3063280052608914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,16384,0.0974044468667772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,51200,1.2296711603800456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,12288,0.21681955125596789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,65536,1.2244062423706055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,7168,128,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,12288,0.07606844769583808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,12288,0.2260248925950792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,51200,0.9505866368611654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,10240,0.06483378012975057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,12288,0.23692443635728624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,10240,0.19056888421376547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,10240,0.18013066715664336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,8192,0.053095999691221446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,7168,0.12522933218214247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,8192,0.15108088652292886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,8192,0.15034222602844238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,65536,1.7422160042656794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,6144,0.10867910914950901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,7168,0.04667911264631483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,7168,0.13502132892608643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,5120,0.07359821928872003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,7168,0.13247022363874647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,6144,0.04122844338417053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,16384,0.28916088740030926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,6144,0.11688710583580865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,10240,0.17766666412353516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,4096,0.07413422399097018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,6144,0.11571199364132351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,5120,0.03567288981543647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,3584,0.061305774582756885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,5120,0.09839999675750732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,4096,0.03035911255412632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,5120,0.09938844707277085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,8192,0.14322843816545275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,4096,0.07944177918963961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,3072,0.0514284438557095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,4096,0.08133777644899157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,3584,0.02756799923049079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,2560,0.04427199893527561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,3072,0.025256888733969793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,3584,0.07377155621846516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,3072,0.06196622053782145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,2048,0.035495999786588885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,3072,0.06719466712739733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,2560,0.02276533345381419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,2560,0.05425777700212267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,1536,0.02629599968592326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,2048,0.019894222418467205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,2048,0.044734223021401294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,1024,0.01850222216712104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,1536,0.017163554827372234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,2048,0.04866310954093933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,1536,0.04137511054674784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,768,0.015136889285511441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,1024,0.01457244488928053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,1024,0.02767466670937008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,1024,0.03302399979697333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,768,0.013447110851605734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,768,0.023360000716315374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,768,0.027460444304678176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,3584,0.07087466451856825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,256,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,512,0.018562666243977018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,512,0.012424888710180918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,512,0.023384888966878254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,128,0.007551111280918121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,256,0.011556444068749746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,2560,0.05715822511249118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,64,0.006842666616042455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,256,0.01993333299954732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,6144,128,0.011770666473441653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,32,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,128,0.015862221519152325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,1536,0.03643644518322415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,6144,128,0.015640000502268474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,64,0.01611111064751943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,32,0.016072000066439312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,6144,512,0.011744000017642975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,65536,0.3731128904554579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,65536,0.9405039681328667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,51200,0.6862506866455078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,6144,256,0.01648000048266517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,51200,0.2942177719540066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,16384,0.2608586682213677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,65536,1.2698550754123263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,12288,0.20861244201660156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,16384,0.27599644660949707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,65536,1.0543866687350802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,16384,0.24445422490437826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,12288,0.06718044148551093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,51200,0.7786702050103081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,12288,0.20149600505828857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,10240,0.1772622267405192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,10240,0.06583111153708564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,12288,0.2023235559463501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,8192,0.11999732918209499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,10240,0.16598221990797254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,7168,0.10392622152964275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,8192,0.046448889705869884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,8192,0.14226667086283365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,7168,0.04216266671816508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,6144,0.10205155611038208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,8192,0.13522666030459934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,7168,0.11612978246476914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,7168,0.1172106663386027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,51200,0.9628168741861979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,16384,0.08414666520224677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,6144,0.03656355539957682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,6144,0.10005778074264526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,4096,0.0676471127404107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,6144,0.10133066442277695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,5120,0.03227555420663621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,3584,0.059501330057779946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,5120,0.08471644586986965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,4096,0.027408000495698717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,5120,0.08683911297056411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,4096,0.06833333439297147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,4096,0.07406400309668647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,3584,0.024554666545655992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,3584,0.061831997500525586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,2560,0.04296000136269463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,10240,0.15873777866363525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,3072,0.02237866653336419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,3584,0.0661262207561069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,3072,0.05370311273468865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,2048,0.03273866573969523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,3072,0.0591848889986674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,2560,0.02052622205681271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,2560,0.04717688759167989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,1536,0.026269333230124578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,2048,0.017758222089873422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,2560,0.05050488975312975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,5120,0.0733920004632738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,1024,0.019306666321224637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,2048,0.044418666097852916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,1536,0.015620445211728414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,1536,0.03170933326085409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,768,0.015767999821239047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,1536,0.036172442966037326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,1024,0.023983110984166462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,1024,0.028594666057162817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,3072,0.048600001467598804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,512,0.011360000405046674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,768,0.012206222448084088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,768,0.020491555333137512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,256,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,768,0.02473955518669552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,512,0.011697777443461947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,512,0.016539555456903245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,128,0.0069884442620807225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,512,0.02102222210831112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,256,0.011180444724029966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,256,0.014813333749771118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,64,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,256,0.019156444403860304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,128,0.01072355525361167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,5120,32,0.006228444476922353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,128,0.014145778285132514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,2048,0.0391173329618242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,5120,128,0.015017777681350708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,64,0.014132445057233175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,5120,32,0.014092443717850579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,5120,1024,0.013070222404268054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,65536,0.5907653172810873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,51200,0.48954041798909503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,16384,0.17928266525268555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,51200,0.23744887775845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,65536,0.9985111024644641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,51200,0.7906079822116427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,16384,0.0726648900243971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,12288,0.1365440024269952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,65536,0.8938710954454211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,16384,0.23628976609971789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,51200,0.6682426664564344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,10240,0.12188711431291367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,16384,0.2080399990081787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,12288,0.05702310800552368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,12288,0.16567822297414145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,8192,0.09829422500398423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,12288,0.16090577178531223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,10240,0.05594578054216173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,10240,0.13858311706119114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,7168,0.07907644245359632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,8192,0.03898577888806661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,10240,0.14646932813856336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,8192,0.12728177176581487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,6144,0.07274133629269071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,8192,0.11178933249579535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,7168,0.03529689047071669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,7168,0.10312266482247247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,5120,0.05224177903599209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,6144,0.031042665243148804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,7168,0.10055289003584121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,6144,0.0864222248395284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,4096,0.04807733496030172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,6144,0.08960888783137004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,5120,0.027377777629428442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,5120,0.07151377863354154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,3584,0.04051999913321601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,4096,0.023677332533730402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,5120,0.07556088765462239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,3072,0.035573333501815796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,4096,0.05825422207514445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,3584,0.022315555148654517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,3584,0.0518853333261278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,2560,0.03015822172164917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,3072,0.01993600030740102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,3584,0.05744533406363594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,3072,0.045904888047112354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,2048,0.024077332682079736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,3072,0.050287998384899564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,2560,0.0181057784292433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,65536,0.3064897855122884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,1536,0.019096889429622226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,2560,0.04456355505519443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,2048,0.016032889485359192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,1024,0.013790222505728403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,2048,0.038260443343056574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,1536,0.01385600037044949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,1536,0.027077333794699773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,768,0.011711999773979187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,1536,0.03226133187611898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,1024,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,1024,0.021065778202480737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,1024,0.025755556093321905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,512,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,768,0.017631111873520743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,768,0.021617778473430212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,4096,0.0629440016216702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,256,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,512,0.014254222313563028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,512,0.010439111126793755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,512,0.019066666563351948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,128,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,256,0.009959999885823991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,256,0.012888888518015543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,2560,0.03984977801640829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,64,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,256,0.017160000072585214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,4096,32,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,2048,0.033482667472627424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,128,0.01239466667175293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,4096,128,0.014126222994592456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,64,0.012433778080675336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,4096,32,0.012114666402339935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,768,0.011077333655622272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,65536,0.6565057966444228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,65536,0.2917368941836887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,51200,0.4383280012342665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,16384,0.16064000129699707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,65536,0.8730426894293891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,51200,0.23166934649149576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,4096,128,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,51200,0.6955831315782336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,65536,0.8405769136216906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,12288,0.13195289505852595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,16384,0.19621333811018202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,16384,0.21580444441901314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,10240,0.1185235579808553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,12288,0.05452088846100701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,51200,0.6421164406670464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,12288,0.15008978048960367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,10240,0.04658222198486328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,12288,0.18088800377315947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,10240,0.12420445018344456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,10240,0.1347279946009318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,7168,0.07976444562276204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,8192,0.09938133425182766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,6144,0.06904000043869019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,8192,0.1090737779935201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,7168,0.03433155682351854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,7168,0.08773689137564765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,5120,0.05684622128804525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,6144,0.03042666779624091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,6144,0.07565689086914062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,16384,0.07270666625764635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,4096,0.04666577776273092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,6144,0.08515822225146824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,5120,0.026518222358491685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,5120,0.06460977925194634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,3584,0.03942577706442939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,5120,0.07413688633177015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,4096,0.02274844381544325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,4096,0.05258311165703667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,3072,0.03346755438380771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,8192,0.09112977981567383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,3584,0.020797333783573575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,3584,0.04734399914741516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,2560,0.028584887584050495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,3584,0.056221332814958364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,3072,0.019021333919631112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,3072,0.04177688890033298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,2048,0.023610666394233704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,3072,0.050159109963311084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,2560,0.01702311138312022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,2560,0.03654933306905959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,1536,0.019527999891175162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,7168,0.09790133105383979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,2048,0.030680888228946265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,2048,0.03757155603832669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,1024,0.014828445182906257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,1536,0.013566222455766467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,1536,0.02460355559984843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,1536,0.030127998855378892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,768,0.011064888702498542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,1024,0.011732444167137146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,1024,0.01908266709910499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,4096,0.06164089176389906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,512,0.008912000391218398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,768,0.011062222222487131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,768,0.016247999336984422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,768,0.02165422174665663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,8192,0.03828355669975281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,256,0.006565333240562015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,512,0.010262221925788457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,512,0.013289777768982781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,128,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,512,0.018536888890796237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,256,0.009684444301658207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,256,0.012088889049159156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,2560,0.04370133413208855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,64,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,256,0.016493333710564505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,128,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,128,0.011384000380833944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,128,0.012784889174832238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,64,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3584,32,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3584,1024,0.024361777636739943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,65536,0.5546808772616917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,65536,0.254040002822876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,51200,0.36309154828389484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,65536,0.7875982390509711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3584,2048,0.015244444211324057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3584,32,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,16384,0.14159022437201604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,51200,0.20542400413089326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,51200,0.659079975552029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,16384,0.06798488563961454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,51200,0.5252719985114204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,16384,0.17139111624823677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,10240,0.0957875582906935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,16384,0.17218400372399223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,12288,0.056003557311164014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,12288,0.1343137820561727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,8192,0.07896444532606336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,10240,0.04855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,10240,0.10839288764529759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,10240,0.11424267292022705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,7168,0.0671164459652371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,8192,0.08967555231518215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,8192,0.035813334915373064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,8192,0.09325333436330159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,6144,0.05673066775004069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,65536,0.6642888916863335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,7168,0.07850755585564508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,5120,0.04451377855406868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,12288,0.10961421993043687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,7168,0.08290311363008288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,6144,0.02924799919128418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,6144,0.06696177853478326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,4096,0.03666311171319749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,6144,0.07281955745485094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,5120,0.025341333614455328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,3584,0.033180445432662964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,5120,0.05792444282107883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,4096,0.021677333447668288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,4096,0.0470106667942471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,5120,0.06298577785491943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,3072,0.026447110705905493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,4096,0.05367288986841837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,12288,0.1350124412112766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,3584,0.020052444603708055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,2560,0.022879999544885423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,3584,0.04244533181190491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,3584,0.048467556635538735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,3072,0.017975999249352347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,3072,0.03772088885307312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,2560,0.016671111186345417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,3072,0.04269955555597941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,2560,0.03256977929009332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,1536,0.015079999963442484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,2560,0.03809244434038798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,2048,0.014922666880819531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,2048,0.027505778604083594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,7168,0.03252977795071072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,2048,0.03238577644030253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,1536,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,768,0.00926755534278022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,1024,0.010671111444632212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,1024,0.017260443833139207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,1024,0.021407999926143225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,512,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,768,0.015050666199790107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,768,0.01018044435315662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,768,0.01927199959754944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,256,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,512,0.009489778015348647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,512,0.012100444071822695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,512,0.01777777738041348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,128,0.005136888888147143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,256,0.009111111362775167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,2048,0.01925155520439148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,256,0.011051555474599203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,64,0.004527111020353106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,256,0.015311999453438653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,32,0.004592888885074192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,3072,128,0.009300444689061906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,128,0.010681777364677854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,3072,1024,0.010891555911964841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,128,0.012111999922328524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,64,0.01056977775361803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,1536,0.022042666872342426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,3072,32,0.01052266690466139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,3072,1536,0.02699377801683214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,65536,0.43236621220906574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,51200,0.3388355573018392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,65536,0.23932178815205893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,16384,0.12089688248104519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,51200,0.18459910816616484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,12288,0.09949688778983222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,16384,0.05642044544219971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,65536,0.65866576300727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,51200,0.5163644684685601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,16384,0.1501120063993666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,16384,0.170744882689582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,65536,0.657335122426351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,10240,0.08224622408548991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,12288,0.04430933462248909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,12288,0.11097066932254368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,8192,0.06971999671724108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,12288,0.13275378280215794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,10240,0.09418133232328628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,10240,0.11998132864634196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,7168,0.057364443937937416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,8192,0.07658577627605863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,8192,0.03157155712445577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,6144,0.048892445034450956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,8192,0.09163110786014134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,7168,0.028192000256644353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,7168,0.06959199905395508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,7168,0.08119999700122409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,6144,0.025973333252800837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,5120,0.04080533319049411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,6144,0.058839109208848744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,4096,0.03242933419015672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,6144,0.07138755586412218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,5120,0.022328888376553852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,5120,0.05043733451101515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,51200,0.5078070958455404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,3584,0.029024001624849107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,4096,0.019088000059127808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,4096,0.041548444165123835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,3072,0.025174222058720056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,10240,0.0381991101635827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,3584,0.01736000014675988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,3584,0.037552889850404524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,3584,0.04641155401865641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,3072,0.016182222300105624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,2560,0.022466666168636743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,3072,0.03281866510709127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,2048,0.018459555175569322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,3072,0.041801777150895864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,2560,0.01455200049612257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,2560,0.028945777151319716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,1536,0.014753777119848462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,2048,0.013384888569513956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,2048,0.024354666471481323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,1024,0.011406222151385413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,2048,0.03169599837727017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,1536,0.01997155613369412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,1536,0.02678666677739885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,768,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,1024,0.015278221832381355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,1024,0.010393777655230628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,5120,0.062260442309909396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,512,0.008064000142945183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,1024,0.021033777130974665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,768,0.009947555760542551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,4096,0.05208977725770739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,768,0.013410666750537025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,256,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,768,0.018615111708641052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,512,0.009272889130645329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,512,0.011069333387745751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,128,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,512,0.017837334010336135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,256,0.008939555949634975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,256,0.010103111465771994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,64,0.004190222256713443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2560,32,0.004400889078776042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,256,0.014816888504558139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,128,0.008749333520730337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,128,0.009661333428488838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,2560,0.03670755690998501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,64,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2560,32,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2560,1536,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,65536,0.37241244316101074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,51200,0.24919732411702475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,16384,0.08868977758619521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,51200,0.1394711070590549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,65536,0.5246808793809679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,65536,0.5177493095397949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,51200,0.4028613302442763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,16384,0.0454053348965115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,12288,0.07590044207043119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,16384,0.12855999999576145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,10240,0.06051377455393473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,16384,0.14406666490766737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,12288,0.03637511200375027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,12288,0.09579555855857001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,8192,0.047522667381498546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,10240,0.03153689040078057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,12288,0.10559111171298557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,10240,0.08120089107089572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,10240,0.09091199768914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,8192,0.02442933287885454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,8192,0.06598577896753947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,65536,0.17363999949561226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,8192,0.07550666729609172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,6144,0.03611288799179925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,7168,0.02216888964176178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,7168,0.05881155861748589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,7168,0.06625066863165961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,5120,0.030051555898454454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,6144,0.050943109724256724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,6144,0.05867111020618015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,51200,0.4029226568010118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,4096,0.024396444360415142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,5120,0.04315911067856682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,5120,0.017849778135617573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,5120,0.051606221331490405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,3584,0.02186311119132572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2560,128,0.01197866681549284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,4096,0.015473778049151102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,4096,0.0358533329433865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,3072,0.019298666053348117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,4096,0.042790220843421094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,7168,0.04483644498719109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,3584,0.014444443914625378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,2560,0.0162773331006368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,3584,0.038478222158220075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,3072,0.013415999710559845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,3072,0.028607110182444256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,2048,0.014118222726715935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,3072,0.034318221939934626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,2560,0.012307555311255984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,2560,0.02481777800454034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,1536,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,6144,0.02027555637889438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,2560,0.030576000610987347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,2048,0.021149333980348375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,2048,0.02602488961484697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,1536,0.01016622202263938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,1536,0.01737866633468204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,768,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,1024,0.013719999955760108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,1024,0.009353777600659264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,1024,0.01793599956565433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,768,0.011765333513418833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,768,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,768,0.016513778103722464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,3584,0.03242311212751601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,256,0.004538666870858935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,512,0.009951111343171861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,512,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,512,0.014509333504570855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,128,0.003939555750952827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,256,0.0081377774477005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,256,0.008970666262838576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,256,0.013392888837390475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,2048,0.01108266661564509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,64,0.003485333174467087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,1024,0.00867377801073922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,2048,128,0.008010666403505538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,32,0.003678222083383136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,128,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,64,0.008349333372380998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,32,0.008742222355471717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,2048,1536,0.02198311189810435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,2048,512,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,51200,0.24289067586263022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,65536,0.4154044522179498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,16384,0.07391377952363756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,51200,0.32992622587415904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,65536,0.5028568903605143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,16384,0.10626311434639825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,16384,0.04417777723736233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,51200,0.39759466383192277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,2048,128,0.008737777670224508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,16384,0.1359528832965427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,10240,0.04779111014472114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,12288,0.034485333495669894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,12288,0.07947555515501234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,65536,0.2943440013461643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,12288,0.1042791141404046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,8192,0.039284444517559476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,10240,0.031261334816614784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,65536,0.16057066122690836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,10240,0.08823021915223862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,7168,0.031481779283947416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,8192,0.02375288969940609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,8192,0.07351111041174994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,51200,0.12945866584777832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,6144,0.02699288891421424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,7168,0.021819555097156104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,7168,0.04899200134807163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,5120,0.02274399995803833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,7168,0.0657048887676663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,6144,0.01954844428433312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,6144,0.04242222176657783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,12288,0.05748533540301853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,4096,0.019425777925385367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,6144,0.057722667853037514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,5120,0.017272000511487324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,5120,0.03654488921165466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,3584,0.01718933383623759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,5120,0.049865778949525624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,4096,0.01536355581548479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,3072,0.015711999601787992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,3584,0.027377777629428442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,3584,0.01443377799457974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,4096,0.04189599884880913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,3584,0.03797333439191183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,10240,0.0674720009167989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,2560,0.013094222380055321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,3072,0.02460977766248915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,3072,0.01308444473478529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,2048,0.011074666347768573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,8192,0.05514311128192478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,2560,0.01205599970287747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,2560,0.021352888809310064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,1536,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,2560,0.02959644463327196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,2048,0.01089777797460556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,2048,0.018192888961897958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,1024,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,2048,0.025565332836574976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,1536,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,1536,0.014974221587181091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,768,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,1536,0.021504888931910198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,1024,0.009306666751702627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,1024,0.017838221457269456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,768,0.008977777428097194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,512,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,4096,0.03024266825781928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,768,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,256,0.003937777959638172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,768,0.016232000456915963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,512,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,512,0.00886577781703737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,128,0.003499555504984326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,256,0.00793333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,256,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,64,0.003212444484233856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,3072,0.03342399994532267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,256,0.012813333008024426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1536,32,0.0032133333798911837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1536,128,0.008016000191370646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,128,0.007928888830873702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,128,0.010402667025725046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,64,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,65536,0.17780800660451254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,32,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,51200,0.14698755741119385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,65536,0.11700622240702312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1536,1024,0.012110222544935016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,16384,0.05526044302516513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,65536,0.3234906726413303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,51200,0.09283822112613255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,51200,0.25849511888292104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,12288,0.039340443081325956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,65536,0.36624799834357363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,16384,0.032088001569112144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,16384,0.08405511246787177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,10240,0.03378666771782769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,51200,0.28584088219536674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,12288,0.025242666403452556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,8192,0.02535111043188307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,12288,0.06343911091486613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,12288,0.07769599888059828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,10240,0.05310844381650289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,7168,0.021934222843911912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,10240,0.06609955761167738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,8192,0.017320000463061862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,8192,0.04422577884462145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,6144,0.019308444526460435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,8192,0.055340445703930326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,7168,0.015856888559129503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,7168,0.03917866614129808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,5120,0.016759110821617972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,7168,0.04918044474389818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,6144,0.01460088955031501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,6144,0.034352001216676496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,4096,0.013775111072593264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,6144,0.043486220969094165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,5120,0.013167999684810638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,5120,0.029337777031792536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,3584,0.012233778006500669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,5120,0.0376800000667572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,4096,0.01183733344078064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,4096,0.02445777753988902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,3072,0.01072177787621816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,16384,0.1013573341899448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,4096,0.03211199906137254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,3584,0.011219555305110084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1536,512,0.01475288967291514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,3584,0.022106667359670002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,3072,0.010681777364677854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,3072,0.0199626667631997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,2048,0.007954667011896769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,2560,0.010082667072614035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,2560,0.022907555103302002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,2048,0.015229332778188916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,2048,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,1536,0.006871999965773688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,2048,0.020305777589480083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,1536,0.008963555925422246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,1024,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,1536,0.012805333567990197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,1536,0.01757600075668759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,10240,0.02270488937695821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,3584,0.028463999430338543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,768,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,1024,0.00830844458606508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,3072,0.026611554953787062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,768,0.009069333473841349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,768,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,512,0.003859555555714501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,768,0.013777777552604675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,512,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,512,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,512,0.01276177747382058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,256,0.006992888947327931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,1024,0.010067555639478896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,256,0.011662222445011139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,256,0.007277333074145847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,1024,0.014998222390810648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,2560,0.01738133364253574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,128,0.007000889215204451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,64,0.0028151110228565005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,1024,128,0.007270221908887227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,32,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,1024,128,0.008986666798591614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,65536,0.16225155194600424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,64,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,1024,32,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,256,0.0032435556252797446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,51200,0.1357253392537435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,65536,0.276180452770657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,65536,0.11382310920291477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,16384,0.04422755704985725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,51200,0.0908542209201389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,2560,0.009338666995366415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,51200,0.21958933936225045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,65536,0.3649262322319879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,12288,0.031114667654037476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,16384,0.030994666947258845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,16384,0.07279022534688313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,10240,0.025570667452282373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,51200,0.2834693325890435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,16384,0.10015110837088691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,12288,0.02421155571937561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,8192,0.021838222940762837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,12288,0.05542044507132637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,10240,0.022384888596004907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,12288,0.07612533039516874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,7168,0.017719111508793302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,10240,0.04667911264631483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,8192,0.017305778132544625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,10240,0.06527288754781087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,6144,0.015483554866578845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,8192,0.03870577613512675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,8192,0.0543431109852261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,7168,0.01567555632856157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,5120,0.013152889079517789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,7168,0.034487111700905695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,6144,0.014685332775115967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,7168,0.048657778236601085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,4096,0.011617778076065911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,6144,0.030527111556794908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,5120,0.013111111190583972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,6144,0.043192889955308705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,3584,0.00998577806684706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,5120,0.026241777671708003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,4096,0.01183555523554484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,5120,0.036193778117497764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,3072,0.008960000342792934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,4096,0.021705778108702764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,3584,0.011145778000354767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,2560,0.007829333345095316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,1024,128,0.0028773333049482773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,3584,0.019665777683258057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,3072,0.010693333215183683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,3584,0.02789777848455641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,2048,0.007062222394678328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,3072,0.017880888448821176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,2560,0.010429333481523724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,3072,0.02548533346917894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,1536,0.0057653333577844836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,2048,0.009687110781669617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,2560,0.022673777408070032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,2048,0.013610666824711693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,1024,0.004519111166397731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,2048,0.02001688877741496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,1536,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,768,0.00416355538699362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,1536,0.01720888912677765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,1024,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,1024,0.009301332963837517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,512,0.0034808889031410217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,1024,0.014456000592973499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,768,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,768,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,256,0.0029351111087534162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,768,0.013435555828942193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,512,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,512,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,128,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,512,0.012392889294359418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,4096,0.0306986669699351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,256,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,256,0.011175110936164856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,128,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,64,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,128,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,768,32,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,2560,0.01551644504070282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,768,128,0.009003555609120263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,65536,0.12410844696892633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,64,0.006098666538794835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,1536,0.008852444589138031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,768,32,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,51200,0.09260800149705674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,65536,0.11094578107198079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,16384,0.03107288810941908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,51200,0.08916622400283813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,51200,0.1845475567711724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,65536,0.3486951192220052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,12288,0.022594667143291895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,16384,0.06120710902743869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,16384,0.03020355436537001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,768,256,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,10240,0.0184568895234002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,12288,0.023522666758961145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,12288,0.04717422193951077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,12288,0.07406310902701484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,10240,0.021241777473025854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,10240,0.04002844293912252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,10240,0.06282044119305082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,8192,0.016533333394262526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,8192,0.03324800067477756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,65536,0.23309866587320963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,8192,0.052631109952926636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,6144,0.011692444483439127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,7168,0.029976000388463337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,7168,0.01514844430817498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,7168,0.0471111105548011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,51200,0.27452267540825737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,5120,0.010316444767846001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,6144,0.014056889547242058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,6144,0.04015555646684434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,5120,0.012614222036467658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,4096,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,5120,0.02278222143650055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,5120,0.035135110219319664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,3584,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,8192,0.014527999692493014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,4096,0.011400000088744693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,4096,0.01891288823551602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,3072,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,4096,0.029912001556820337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,3584,0.010853333605660332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,7168,0.013078221844302284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,3584,0.017384888397322763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,2560,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,3584,0.027130666706297133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,3072,0.015615110596021017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,2048,0.005392888767851724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,3072,0.02445777753988902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,2560,0.009694221946928237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,2560,0.0222053329149882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,1536,0.004644444419278039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,2048,0.009127111070685917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,2048,0.012105777859687805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,6144,0.026439110438028973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,1024,0.0038577777643998465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,2048,0.019595555133289762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,16384,0.09666399823294745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,1536,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,1536,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,768,0.00350577798154619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,1536,0.016896888613700867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,1024,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,1024,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,512,0.003136000078585413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,1024,0.014441778262456259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,768,0.0076657773719893555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,768,0.007781333393520779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,256,0.0027431110954946945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,768,0.013564444250530668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,512,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,512,0.006647999915811751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,128,0.0025271111064487034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,512,0.012375999655988483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,256,0.006990222467316522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,3072,0.01033155537313885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,256,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,32,0.002510222295920054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,512,128,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,128,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,2560,0.013870221873124441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,128,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,65536,0.07641955878999498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,64,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,51200,0.06139466497633192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,512,32,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,65536,0.11049333545896743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,65536,0.19047021865844727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,51200,0.08867733346091376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,51200,0.14903377162085638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,65536,0.34244267145792645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,51200,0.26984087626139325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,16384,0.0505600008699629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,12288,0.0158924443854226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,16384,0.029383109675513372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,16384,0.09535111321343316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,10240,0.015371556083361307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,12288,0.022671111755900916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,512,64,0.0022960000981887183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,12288,0.03872177667087979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,8192,0.01292888902955585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,512,256,0.011062222222487131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,12288,0.07246310843361749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,10240,0.0209840006298489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,7168,0.011191111471917896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,10240,0.033257779147889875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,8192,0.016302223006884258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,10240,0.0621066689491272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,8192,0.027684445182482403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,6144,0.010016889207892949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,8192,0.05127822359402975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,7168,0.01514933341079288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,7168,0.024728889266649883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,16384,0.021924444370799597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,7168,0.04557866520351834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,6144,0.013801777528391944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,6144,0.02199911077817281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,4096,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,6144,0.04018044471740723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,5120,0.01255644443962309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,3584,0.006865777903132969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,5120,0.019292443990707397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,5120,0.03410044312477112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,4096,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,3072,0.006044444524579578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,4096,0.01642933322323693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,4096,0.029366221692827012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,3584,0.010853333605660332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,3584,0.015054222610261707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,2560,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,3584,0.026737777723206416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,3072,0.01036088913679123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,3072,0.013786666923099093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,2048,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,3072,0.02401599950260586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,2560,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,1536,0.003882666842805015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,2560,0.02131555477778117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,2048,0.00903733323017756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,2048,0.011009777585665384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,1024,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,2048,0.01924800044960446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,1536,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,1536,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,768,0.003089777711364958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,1536,0.016373333003785875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,1024,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,1024,0.008263111114501953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,512,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,1024,0.014135110709402295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,768,0.007667555577225155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,256,0.002492444382773505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,768,0.007205333146784041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,512,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,512,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,768,0.012716444830099741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,128,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,512,0.01149866647190518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,256,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,256,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,64,0.0021537778278191886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,32,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,256,0.010762666662534079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,128,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,2560,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,65536,0.054808888170454234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,64,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,256,32,0.005304000029961268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,51200,0.044808000326156616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,65536,0.11007289091746013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,256,5120,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,65536,0.1894035604265001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,16384,0.01200977795653873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,51200,0.08862133158577813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,51200,0.14826044771406385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,65536,0.3389688862694635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,12288,0.012863111164834766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,16384,0.05021422108014425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,51200,0.267631106906467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,16384,0.09398844507005479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,10240,0.011681777735551199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,12288,0.022410665949185688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,12288,0.03877422213554382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,8192,0.01185066666867998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,12288,0.07169155279795329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,10240,0.020759999752044678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,256,128,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,10240,0.0326257778538598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,7168,0.010823999842007955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,10240,0.06067733632193672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,256,128,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,8192,0.016255110502243042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,6144,0.00960000024901496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,8192,0.0270506673389011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,7168,0.015185778339703878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,8192,0.05017244484689501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,5120,0.00831200016869439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,7168,0.04418399930000305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,6144,0.01389955480893453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,6144,0.03952088952064514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,4096,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,5120,0.012532444463835822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,5120,0.018996444013383653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,16384,0.029321776496039495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,3584,0.006503110958470239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,5120,0.033668445216284856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,4096,0.011424000064531961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,4096,0.016207110550668504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,3072,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,4096,0.02884888980123732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,3584,0.01073155552148819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,3584,0.014815110299322339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,2560,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,3584,0.026130666335423786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,3072,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,3072,0.01348977784315745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,2048,0.004509333521127701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,2560,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,2560,0.012063110868136087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,7168,0.024100444383091394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,2560,0.020992888344658747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,1536,0.003866666720973121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,2048,0.009008888569143083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,2048,0.010693333215183683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,6144,0.021472889516088698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,1024,0.003160888950030009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,2048,0.01889511115021176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,1536,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,768,0.0028782222006056043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,1536,0.016197333733240765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,1024,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,1024,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,512,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,1024,0.013748444616794586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,768,0.007627555893527136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,768,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,256,0.002471999989615546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,768,0.012423999607563019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,512,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,512,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,128,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,512,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,256,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,256,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,128,32,0.002201777779393726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,3072,0.023704888092146978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2304,128,128,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,128,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,65536,0.05261422197024027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,51200,0.04186755418777466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,128,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,16384,0.01516266663869222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,32,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,12288,0.012497777740160624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,10240,0.011063999599880643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,65536,0.18832355075412324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,1536,0.009180444810125563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,51200,0.14807021617889404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,16384,0.05003644360436333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,7168,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,12288,0.03789866632885403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,6144,0.008449777960777283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,10240,0.03241066800223456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,5120,0.007840889195601145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,4096,0.00702044450574451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,8192,0.026757333013746474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,7168,0.02401599950260586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,3584,0.006855999843941794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,6144,0.021566222111384075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,5120,0.01888088881969452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,2560,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,4096,0.016185777054892648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,2048,0.004559111264016894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,3584,0.01479200025399526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,1536,0.003863111138343811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,3072,0.013452444639470843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2304,128,256,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,1024,0.0031902222997612427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,2560,0.01204533295498954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,2048,0.01075644459989336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,128,64,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,512,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,256,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,1536,0.009030222064918941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,1024,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,128,0.0018622221218215094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,64,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,768,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,512,0.006435555716355641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,8192,0.009711110757456886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,128,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,65536,0.048002666897243924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,64,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,51200,0.038803554243511625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,16384,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,32,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,12288,0.008386666576067606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,3072,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,10240,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,65536,0.18858044677310518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,16384,0.04997333221965366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,8192,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,7168,0.00666933341158761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,12288,0.03775644302368164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,6144,0.006520000182920032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,10240,0.032199111249711775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,64,768,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,5120,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,8192,0.026782222919993933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,4096,0.01073333372672399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,7168,0.02405599918630388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,3584,0.010081777969996134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,3072,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,6144,0.021425777011447485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,5120,0.018889778190188937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,2560,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,2048,0.0041946665280395085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,4096,0.016183111402723525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,1536,0.0038408889538711975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,3072,0.01348355578051673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,1024,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,2560,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,768,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,2048,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,512,0.002532444480392668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,256,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,1536,0.00906844437122345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,128,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,1024,0.007763555480374231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,768,0.0069919998447100324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,64,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,512,0.005980444451173146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,256,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,128,0.005051555732885997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,64,0.005064888960785336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,32,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,64,256,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,51200,0.14780622058444554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2304,32,3584,0.014831110835075378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2304,32,32,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,16384,2.737061394585503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,16384,3.0135805341932507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,16384,0.8925626542833117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,12288,2.048118167453342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,12288,0.6742533577813042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,51200,3.2523964775933156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,12288,2.2158586714002824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,10240,0.5529155731201172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,10240,1.705018679300944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,8192,0.4541529019673665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,10240,1.7771191067165797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,8192,1.4783617655436199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,7168,0.3999484380086263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,8192,1.4359928766886394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,6144,1.1123119990030925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,6144,0.3809013366699219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,7168,1.1706248389350045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,6144,1.0093982484605577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,5120,0.29799644152323407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,5120,1.0114195081922743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,7168,1.2642275492350261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,4096,0.6388826900058323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,4096,0.24700445599026152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,4096,0.650711112552219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,5120,0.8607778019375272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,3584,0.23178932401869032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,3584,0.6805040041605631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,51200,8.451245625813803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,3072,0.194794668091668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,3584,0.5705848799811469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,3072,0.6099813249376085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,2560,0.4283528857760959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,3072,0.4901253382364909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,2048,0.3906595442030165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,2048,0.17456177870432535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,2560,0.4579635726081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,1536,0.25939999686347115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,2048,0.3416213459438748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,2560,0.16659555170271131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,1536,0.1411084466510349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,768,0.1333262258105808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,1024,0.20687733756171334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,1536,0.28184178140428334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,1024,0.10227644443511963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,512,0.10244266854392158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,768,0.09027910894817776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,256,0.08535555336210464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,512,0.07402577665117052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,768,0.16666666666666666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,128,0.05854044357935587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,512,0.15543822447458902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,256,0.1257022221883138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,128,0.0638986693488227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,32,0.052948445081710815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,64,0.12153689066569011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,32,0.12106488810645209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,1024,0.20738311608632407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,65536,256,0.07016088565190633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,65536,64,0.053759998745388456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,128,0.11905244986216228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,65536,51200,11.755519443088106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,16384,2.0826310051812067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,16384,2.342807133992513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,51200,2.5780116187201605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,16384,0.6833928955925835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,65536,3.2670800950792103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,12288,0.5863519774542915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,12288,1.5001004536946614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,12288,1.6999297671847875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,10240,0.43298665682474774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,10240,1.5658310784233942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,51200,6.336323632134332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,8192,0.3506462309095595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,10240,1.4462116029527452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,8192,1.0516168806287978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,8192,1.0929910871717665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,7168,0.3168959882524279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,7168,0.9814817640516492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,65536,8.29690466986762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,6144,0.7636071311102973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,6144,0.2749679883321126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,7168,0.9519689348008898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,5120,0.7430444293551974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,6144,0.8123235702514648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,5120,0.22966400782267252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,4096,0.524944888220893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,5120,0.6819030973646375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,3584,0.39533244238959414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,3584,0.18227822250790068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,4096,0.19343199994828966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,3072,0.3429555628034804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,3584,0.4476773473951552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,3072,0.1536897818247477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,2560,0.33230222596062553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,3072,0.3984053399827745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,2560,0.1325964397854275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,2048,0.25937244627210826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,2560,0.35688087675306535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,2048,0.11401689052581787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,1536,0.19811377260420057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,51200,9.619202507866754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,2048,0.305239995320638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,1024,0.15521332952711317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,1536,0.09153244230482314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,1024,0.0743866695298089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,768,0.10967466566297744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,512,0.08427644438213772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,1024,0.17102132903205025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,768,0.0657048887676663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,1536,0.25597866376241046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,256,0.0573564436700609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,512,0.06078755855560303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,128,0.04974666568968031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,512,0.11305422253078884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,256,0.054732445213529796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,64,0.040431999497943454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,51200,32,0.04065333472357856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,128,0.09525955385631985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,64,0.09722044732835557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,32,0.09608533647325303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,4096,0.5185315344068739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,65536,12.100059509277344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,768,0.142922666337755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,51200,128,0.052054223087098866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,65536,0.9925484127468533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,16384,0.6844862302144369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,51200,0.7609297964307996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,16384,0.25862667295667857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,51200,256,0.10071111387676662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,51200,2.085071987575955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,65536,2.6647733052571616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,12288,0.524837334950765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,16384,0.814851548936632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,12288,0.1785200039545695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,10240,0.4236444367302789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,10240,0.18581510914696586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,12288,0.5335795614454482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,8192,0.3449813259972467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,7168,0.29545866118537056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,10240,0.4504399829440647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,8192,0.12088267008463542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,7168,0.303184880150689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,7168,0.10918666919072469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,65536,3.5227385626898875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,6144,0.25275111198425293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,6144,0.09365422195858425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,51200,2.9213538699679904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,5120,0.21576444307963052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,5120,0.08175288968616061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,4096,0.17038844691382515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,3584,0.1590986649195353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,4096,0.17880711290571424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,5120,0.21704444620344374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,3584,0.0616168909602695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,3584,0.1688586605919732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,2560,0.11135021845499675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,3072,0.05417866508165995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,3072,0.13212177488538954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,2048,0.08725155724419488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,2560,0.04873866505093045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,2560,0.11561511622534858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,1536,0.06695022185643514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,2048,0.04292177822854784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,2048,0.0933484435081482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,6144,0.26312798923916286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,1024,0.04599555664592319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,1536,0.033738666110568576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,768,0.039091553952958845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,1536,0.07560266388787164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,4096,0.06635022163391113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,1024,0.027264000640975103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,512,0.027541332774692114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,1024,0.05663910839292738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,768,0.02516355613867442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,8192,0.34858134057786727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,3072,0.11476888921525742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,768,0.04546844296985202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,256,0.019527999891175162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,128,0.01668355531162686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,512,0.03818133473396301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,64,0.014027555783589682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,256,0.021074665917290583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,16384,32,0.014429334137174817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,256,0.03416622347301907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,128,0.020975111259354487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,128,0.032440000110202365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,64,0.032440000110202365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,16384,32,0.032299555010265775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,16384,512,0.023443554838498432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,51200,0.5784319771660699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,65536,0.744547579023573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,16384,0.4476693471272786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,16384,0.18001511361863878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,51200,1.5169430838690863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,12288,0.4819128778245714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,16384,0.5438577863905165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,12288,0.13723377386728922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,65536,2.010380427042643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,12288,0.4029857847425673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,10240,0.32938043276468915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,10240,0.11348266071743435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,8192,0.26840088102552623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,8192,0.09548088577058579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,7168,0.25994133949279785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,10240,0.4753084182739258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,7168,0.09468888574176365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,51200,2.177147547403971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,6144,0.18626933627658418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,8192,0.286519103580051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,7168,0.22693067126803926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,5120,0.17624799410502115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,4096,0.11921421686808269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,5120,0.06522755490409003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,65536,2.7937136756049266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,3584,0.10147555669148763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,5120,0.17533066537645128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,6144,0.26980800098843044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,4096,0.13178843922085232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,3584,0.04996711015701294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,2560,0.08014044496748182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,3072,0.04621600111325582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,3584,0.12104445033603245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,2048,0.06902933120727539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,3072,0.10219377941555446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,2560,0.03714488943417867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,1536,0.05232888791296217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,2560,0.08980977535247803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,1024,0.03487022386656867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,1536,0.026352889007992212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,1536,0.0584391090604994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,768,0.028937776883443195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,1024,0.02179199953873952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,1024,0.04482488830884298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,512,0.020040000478426617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,6144,0.09321155813005234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,768,0.02020355562369029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,256,0.01570133368174235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,768,0.03636799918280707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,512,0.018512000640233357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,4096,0.05795911285612318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,128,0.013086222112178802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,512,0.029342220889197454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,3072,0.08691377772225274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,256,0.01685599982738495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,64,0.011042666931947073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,256,0.026111111044883728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,12288,32,0.011377777490350934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,128,0.01680266691578759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,128,0.025370667378107708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,64,0.02520177761713664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,32,0.025091555383470323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,12288,2048,0.03365155392222934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,12288,2048,0.07288799683252971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,16384,0.3694133228725857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,65536,0.6322595808241103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,51200,0.5004888640509711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,16384,0.18796000215742323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,51200,1.2193280325995552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,12288,0.33161155382792157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,16384,0.4480942090352376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,65536,1.623642709520128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,10240,0.2752755482991536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,10240,0.1037564476331075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,8192,0.22214933236440024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,12288,0.4483911196390788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,51200,1.706877284579807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,7168,0.19537066088782418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,8192,0.23070311546325684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,6144,0.1632711092631022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,7168,0.20724088615841338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,65536,2.6565803951687283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,5120,0.13352266947428384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,6144,0.06588977575302124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,6144,0.16586221588982478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,4096,0.10739822520150079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,5120,0.05352977911631266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,5120,0.17779110537634957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,12288,0.12386043866475423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,3584,0.10135644674301147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,4096,0.04485866758558485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,3072,0.07889333036210802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,4096,0.11234488752153184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,3584,0.039994666973749794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,10240,0.272598213619656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,2560,0.06369155645370483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,3584,0.10407377613915338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,3072,0.0356391105386946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,2560,0.032124443186653986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,3072,0.08914222319920857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,8192,0.0975688894589742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,1536,0.041788445578681103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,2048,0.0277137772904502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,2560,0.07557155688603719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,1024,0.02882933285501268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,768,0.021184888150956895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,2048,0.06267911195755005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,1536,0.022895110978020564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,7168,0.07716000080108643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,1024,0.019501333435376484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,512,0.016864889197879367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,1536,0.049955556790033974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,256,0.012484444512261285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,1024,0.038402666648228966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,768,0.018182223041852314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,128,0.010248888697889116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,512,0.01608800060219235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,768,0.0314284430609809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,256,0.014813333749771118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,64,0.008965333302815756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,32,0.009016888837019602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,512,0.025401777691311304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,10240,128,0.014846222268210517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,128,0.021664000219768945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,64,0.021703999903466966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,32,0.021673777037196692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,10240,2048,0.05029866761631436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,65536,0.5118657747904459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,51200,0.40379020902845597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,65536,1.2661413616604276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,16384,0.12379911210801865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,10240,256,0.02255200015174018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,16384,0.4147520065307617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,12288,0.2640639940897624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,12288,0.09585244125790066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,12288,0.28251645300123424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,10240,0.08158489068349202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,8192,0.17647822697957358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,10240,0.2452951007419162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,65536,2.0267999437120223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,16384,0.34119023217095273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,8192,0.06503378020392524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,8192,0.18085511525472006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,6144,0.13024000326792398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,7168,0.05785510937372843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,7168,0.19456888569725883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,5120,0.11108799775441487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,6144,0.051311112112469144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,51200,1.2334524790445964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,6144,0.1374222172631158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,4096,0.08886400196287368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,5120,0.04353333181805081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,3584,0.08122755421532525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,5120,0.12639199362860784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,4096,0.036321779092152916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,3072,0.06604977448781331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,4096,0.09310577975379096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,3584,0.033364444971084595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,2560,0.05693422423468696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,3072,0.030635555585225422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,3072,0.07221066951751709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,2560,0.02756266627046797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,2048,0.04495199852519565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,2560,0.062126219272613525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,7168,0.15502311123742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,1536,0.03332533439000448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,2048,0.023753777146339417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,1024,0.024587555064095393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,2048,0.052057776186201304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,1536,0.020139555136362713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,768,0.017768889665603638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,1536,0.04140355520778232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,1024,0.016883555385801528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,512,0.014125333891974555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,1024,0.03192977772818671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,768,0.015462223026487561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,256,0.010431999961535135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,768,0.026621333426899377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,512,0.014152000347773233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,128,0.00923111124171151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,51200,1.3674791124131944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,256,0.01311555587583118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,64,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,256,0.019194665882322524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,8192,128,0.01311644415060679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,32,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,64,0.018566222654448617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,32,0.01826844447188907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,65536,0.49515289730495876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,512,0.021565333008766174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,65536,1.203701337178548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,3584,0.08213511440489027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,8192,128,0.01827022267712487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,16384,0.11115911271837021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,16384,0.3157911035749647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,51200,0.3796524471706814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,51200,0.7668586836920844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,16384,0.31965245140923393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,12288,0.24007911152309844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,12288,0.0843226644727919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,10240,0.1827039983537462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,65536,1.8474239773220484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,8192,0.15682310528225368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,10240,0.07283022006352742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,12288,0.2743368943532308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,10240,0.19819377528296578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,7168,0.13396444585588244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,8192,10240,0.21596711211734346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,6144,0.11073866817686294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,8192,0.17059644063313803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,51200,1.2759360207451713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,7168,0.1459253364139133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,6144,0.04534844557444254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,4096,0.0816355546315511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,5120,0.04246222310596042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,5120,0.10424533155229355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,3584,0.07410399781333075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,4096,0.0830391115612454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,3072,0.05551911062664456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,3584,0.0321066677570343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,3584,0.07369244760937162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,3072,0.027522666586769953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,2560,0.04284088810284933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,3072,0.06457333432303534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,2048,0.0383200016286638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,8192,0.06777866681416829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,2560,0.0567457808388604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,7168,0.05245777633455065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,1536,0.0284204449918535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,2048,0.02186577849917942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,5120,0.09158755673302545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,2048,0.04653955499331156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,1024,0.019966221517986722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,768,0.0162782222032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,1536,0.018537777993414137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,6144,0.1237306727303399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,1536,0.03750133183267381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,512,0.013400889105266996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,768,0.014391111003028022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,1024,0.028905779123306274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,4096,0.03346133232116699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,256,0.009361777868535783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,512,0.013239111337396832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,768,0.02423111100991567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,128,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,64,0.00721688899728987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,512,0.019319110446506076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,256,0.012423999607563019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,7168,32,0.007585778004593319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,128,0.012422222230169507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,256,0.01720977822939555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,128,0.016879111528396606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,64,0.01682844426896837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,7168,32,0.016582222448454965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,2560,0.025076443950335186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,65536,0.38742756843566895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,16384,0.22090666823916963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,51200,0.31826043128967285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,7168,1024,0.015767999821239047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,51200,0.8359760178460015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,16384,0.09575111336178249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,12288,0.18842221630944145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,16384,0.29146311018202037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,12288,0.07357155614429049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,51200,1.0150470733642578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,12288,0.23829600546095106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,10240,0.06321333514319526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,65536,1.514036496480306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,10240,0.1757191154691908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,7168,0.12666311528947619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,8192,0.05077777637375725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,8192,0.14358666208055285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,6144,0.09746044211917454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,7168,0.05394222338994344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,7168,0.1238222254647149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,5120,0.08887200223075019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,6144,0.03969688879119025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,65536,0.9861413107977973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,6144,0.10806044605043198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,4096,0.061853335963355176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,5120,0.037661333878835045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,3584,0.05212355653444926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,5120,0.09015911155276829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,4096,0.030240890052583482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,3072,0.047855999734666615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,10240,0.15282666683197021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,3584,0.027069333526823256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,2560,0.04517777760823568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,3584,0.06717422273423937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,3072,0.02476177778508928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,2048,0.03424088822470771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,3072,0.05766310956743029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,2560,0.022474666436513264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,1536,0.025956445270114478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,2048,0.019955555597941082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,1024,0.018983110785484314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,2048,0.04164888792567783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,1536,0.017058667209413316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,768,0.01519733336236742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,1536,0.03425600131352743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,1024,0.014447111222479077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,512,0.011429333024554782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,1024,0.025859556264347498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,768,0.013459555804729462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,768,0.021733333667119343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,512,0.012372444073359171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,128,0.007009777757856581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,512,0.01738133364253574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,256,0.01146399974822998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,64,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,256,0.01568000018596649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,32,0.006515555497672822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,4096,0.0728693339559767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,128,0.015173332558737861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,64,0.0151573336786694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,8192,0.14545688364240858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,32,0.01510933372709486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,6144,2560,0.050119111935297646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,65536,0.366575108634101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,6144,256,0.00871288859181934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,51200,0.285041782591078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,51200,0.622238212161594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,65536,0.8346062236362033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,16384,0.2327280044555664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,16384,0.08224622408548991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,12288,0.16596266958448622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,6144,128,0.01144088887506061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,51200,0.8877324528164334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,10240,0.14447200298309326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,12288,0.06592800219853719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,65536,1.2265777587890625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,12288,0.19006400638156465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,10240,0.054493334558275014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,7168,0.09164622094896104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,8192,0.04616266820165846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,10240,0.15668888886769614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,6144,0.08258044719696045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,7168,0.039804445372687444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,8192,0.12058399783240424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,5120,0.06943200031916301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,7168,0.10740888781017727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,6144,0.036044445302751325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,4096,0.05644355879889595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,6144,0.0933519999186198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,3584,0.0492159989145067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,4096,0.026055110825432673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,5120,0.07767822345097859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,3072,0.04201155569818285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,3584,0.02843822373284234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,2560,0.03640177845954895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,3584,0.05605777766969469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,3072,0.021828444467650518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,2048,0.02703999976317088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,2560,0.020038222273190815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,2560,0.04603200157483419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,16384,0.2386444409688314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,1536,0.02094488839308421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,2048,0.017431111799346078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,2048,0.03625955515437656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,8192,0.11191377374860977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,1024,0.015156444576051501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,1536,0.015508444772826301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,768,0.012448000411192576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,1536,0.02976800004641215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,1024,0.01313244468635983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,512,0.00942311104800966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,1024,0.022330666581789654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,768,0.012465777496496836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,256,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,5120,0.031045334206687078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,768,0.01936444474591149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,512,0.011481777661376528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,512,0.015231110983424716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,256,0.010957332948843638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,4096,0.06343289216359456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,64,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,256,0.014019555515713163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,32,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,5120,128,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,128,0.013455111119482251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,3072,0.049695998430252075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,64,0.013439999686347114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,5120,32,0.013396444420019785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,65536,0.29691823323567706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,65536,0.5491066508822972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,51200,0.5271217558119032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,51200,0.23543998930189344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,16384,0.17395555973052979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,12288,0.13018400139278835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,16384,0.08049155606163873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,12288,0.06365244256125556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,5120,128,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,10240,0.10625511407852173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,16384,0.2057653268178304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,65536,0.8930240207248263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,51200,0.6865928967793783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,8192,0.0982817808787028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,12288,0.14723200268215603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,7168,0.08540088600582546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,10240,0.05189955565664503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,6144,0.06398577822579278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,7168,0.03413777881198459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,10240,0.13109866778055826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,6144,0.0318248901102278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,7168,0.10193066464530097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,4096,0.04734933376312256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,6144,0.07944355408350627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,3584,0.03863911165131463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,4096,0.02366133365366194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,4096,0.05530577898025513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,3072,0.03495111068089803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,3584,0.02141777839925554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,3584,0.04848177896605598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,2560,0.02875022093454997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,3072,0.04296711087226868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,2560,0.017495110630989075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,2560,0.037350220812691584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,2048,0.024139554964171514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,8192,0.038372443781958684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,1536,0.01862133377128177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,5120,0.05235822333229912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,2048,0.015503111812803479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,8192,0.10454222228791978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,1024,0.013847110999955071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,2048,0.03137777911292182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,768,0.011358222199810876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,1536,0.013776000175211163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,5120,0.027833776341544256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,1024,0.011956444217099084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,1536,0.025065778030289546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,5120,0.06593688991334704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,1024,0.0196782218085395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,768,0.011012444065676795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,256,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,128,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,768,0.016545777519543965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,512,0.010404444403118556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,64,0.00499911109606425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,256,0.009698666632175446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,32,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,128,0.00998399986161126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,256,0.012103111379676394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,4096,3072,0.019341333044899833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,128,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,64,0.011707555916574268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,32,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,65536,0.28022223048739964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,65536,0.4962755309210883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,51200,0.47391022576226133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,51200,0.22534134652879503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,16384,0.14983644750383165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,12288,0.11546488602956136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,16384,0.08076000213623047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,4096,512,0.00850133349498113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,10240,0.09655733240975274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,16384,0.18502399656507704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,65536,0.8053075472513834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,51200,0.6222346623738607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,12288,0.05484444565243191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,8192,0.07047466437021892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,4096,512,0.013622221847375235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,10240,0.051663110653559365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,7168,0.06085511048634847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,10240,0.1111982266108195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,8192,0.03822933302985297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,6144,0.05266133281919691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,8192,0.09442133373684353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,5120,0.0480968885951572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,6144,0.03071999881002638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,7168,0.07960000303056505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,4096,0.03996710975964864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,6144,0.06957955492867364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,5120,0.02693155573474036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,3584,0.03399111164940728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,5120,0.060327112674713135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,4096,0.022188444932301838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,3072,0.02935911218325297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,4096,0.048226667775048145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,3584,0.0203155560625924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,2560,0.0243848893377516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,3584,0.043663998444875084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,3072,0.01844888925552368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,2048,0.021125333176718816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,3072,0.038663112454944186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,2560,0.016869333055284288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,1536,0.016695110334290397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,2048,0.015071110592948066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,1024,0.012823999755912356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,2048,0.02873777680926853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,12288,0.13801422384050158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,768,0.010784000158309937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,1536,0.02277066641383701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,1024,0.011725333001878528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,512,0.00870666652917862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,768,0.010832889212502373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,7168,0.03459377752410041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,768,0.01529600057337019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,512,0.010211555494202508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,512,0.012607999973826937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,128,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,256,0.009588444398509132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,256,0.011408000356621213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,64,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,128,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,32,0.004972444640265571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,128,0.011041777829329172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,64,0.010851555400424533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,2560,0.033943110042148165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,32,0.011017777853541903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3584,1536,0.013454222016864352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,65536,0.42961955070495605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,65536,0.23994400766160753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,16384,0.13235467010074192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,51200,0.1962062252892388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3584,256,0.007583111524581909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,12288,0.1032231118943956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,16384,0.06478844748602973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,51200,0.5278471310933431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,65536,0.6944230927361382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,10240,0.09670044316185845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,12288,0.050231112374199756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,12288,0.11783377329508464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,10240,0.044030222627851695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,7168,0.06707821951972114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,8192,0.03489777776930068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,10240,0.1014488869243198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,6144,0.052969776921802096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,7168,0.03141866789923774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,7168,0.07209066549936931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,6144,0.028063111835055884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,5120,0.044172445933024086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,6144,0.062295112344953746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3584,1024,0.017920000685585868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,4096,0.034845332304636635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,51200,0.3816853364308675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,3584,0.03191733360290527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,5120,0.024767110745112102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,4096,0.021194666624069214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,5120,0.05314488874541389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,3072,0.027663111686706543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,4096,0.043268445465299815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,2560,0.023733332753181458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,16384,0.15676711665259466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,3584,0.03934933410750495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,8192,0.07689066727956136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,3072,0.034324444002575345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,2560,0.015963556038008798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,1536,0.016524444023768108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,2560,0.03001333276430766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,2048,0.014708444476127625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,1024,0.011894221935007306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,8192,0.08140533500247531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,1536,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,768,0.00964355551534229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,1536,0.020688888099458482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,1024,0.010535111029942831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,512,0.0074755557709270054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,1024,0.016169778174824186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,768,0.010125333236323463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,768,0.013980444934633044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,256,0.005542222410440445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,512,0.00941066692272822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,512,0.011393778026103973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,128,0.004852444347408083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,256,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,3584,0.018486221631368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,256,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,64,0.004412444515360726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,128,0.00902844468752543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,3072,3072,0.01755644381046295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,128,0.010019555687904358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,2048,0.019913777709007263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,64,0.010051555931568146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,32,0.01017599966790941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,3072,2048,0.02557955516709222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,65536,0.21269333362579346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,65536,0.4484586715698242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,51200,0.31155112054612905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,16384,0.11587111155192058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,51200,0.17326577504475912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,12288,0.08611288997862075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,16384,0.055252446068657764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,65536,0.5934649043613011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,16384,0.1598177750905355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,51200,0.45915820863511825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,10240,0.08527555730607773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,8192,0.06779822376039293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,12288,0.10186044375101726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,10240,0.08710400263468425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,3072,32,0.004538666870858935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,8192,0.03048711021741231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,8192,0.07004000080956353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,6144,0.04936000042491489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,7168,0.027127110295825537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,5120,0.04005333450105455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,6144,0.025133333272404138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,7168,0.06254755788379245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,4096,0.03330488999684652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,6144,0.05411466625001696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,5120,0.021652445197105408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,3584,0.02943555514017741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,4096,0.0184560004207823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,5120,0.0468408895863427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,3072,0.02442577812406752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,4096,0.038189331690470375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,3584,0.01700444519519806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,2560,0.02094488839308421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,3584,0.034685333569844566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,3072,0.01574577722284529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,2048,0.018069333500332303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,12288,0.042767110798094005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,2560,0.014333332578341166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,1536,0.013968000809351603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,10240,0.03927466604444716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,2048,0.013781333135233985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,2560,0.02661422226164076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,1024,0.010319111247857412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,7168,0.05736177497439914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,1536,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,1024,0.010348444183667501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,1536,0.01832711034350925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,512,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,1024,0.014506666196717156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,256,0.005112888912359874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,768,0.012786666552225748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,512,0.010402667025725046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,128,0.004552000098758274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,256,0.009713778065310584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,128,0.008678221868144141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,128,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,64,0.0041911109454101985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,32,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,3072,0.030750221676296655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,64,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,32,0.009296000003814697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2560,768,0.008616000413894653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,65536,0.31908710797627765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,768,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,65536,0.16648088561164007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,512,0.009259555902745988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,51200,0.29510755009121364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,16384,0.0937840011384752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,12288,0.07200621896319918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,51200,0.13298043939802381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,65536,0.4927822219000922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,10240,0.06176533301671346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,16384,0.12043910556369358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,12288,0.03466399841838413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,51200,0.4027395513322618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,8192,0.047134220600128174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,12288,0.09027022123336792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,10240,0.032299555010265775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,7168,0.03984977801640829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,10240,0.07623200284110175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,8192,0.024162666665183172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,6144,0.034861332840389676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,7168,0.02276355524857839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,5120,0.029193778832753498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,7168,0.0547635555267334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,6144,0.0196142229768965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,6144,0.04807910985416836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,4096,0.023385778069496155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,5120,0.017260443833139207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,5120,0.040961778826183744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,4096,0.015207999282413058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2560,256,0.008751110898123847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,16384,0.05418489045566983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,3072,0.017816889617178176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,3584,0.014287999934620328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,3584,0.03109688891304864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,2560,0.015593777100245157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,3072,0.013288000391589271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,3072,0.0273991111252043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,2048,0.01316177762216992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,2560,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,8192,0.06412266360388862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,1536,0.010664000279373594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2560,2048,0.02237155536810557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,2048,0.02029244436158074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,2048,0.011117333339320289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,1024,0.008169777691364288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,1536,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,768,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,1024,0.00923999978436364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,1024,0.013582222163677216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,3584,0.019721777902709115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,512,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,768,0.008967999782827165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,4096,0.03432088759210374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,768,0.011382222175598145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,512,0.00834222220712238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,128,0.0038924444880750445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,256,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,256,0.008700444466537898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,64,0.0035413333939181436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,2048,128,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,128,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,32,0.003548444559176763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,2560,0.02386933399571313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,64,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,65536,0.2672293451097277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,32,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,51200,0.23571909798516166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,51200,0.12277155452304417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,16384,0.07430933581458198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,65536,0.39768534236484104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,2048,256,0.004518222063779831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,51200,0.3107937706841363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,12288,0.058913780583275684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,16384,0.04247466723124186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,10240,0.04658311274316576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,16384,0.10056800312466091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,512,0.009661333428488838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,12288,0.033222221665912204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,8192,0.03683999843067593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,7168,0.031686223215527005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,10240,0.029521776570214167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,12288,0.07477511299981011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,6144,0.02738044493728214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,8192,0.022981334063741896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,7168,0.021223111285103693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,8192,0.05151822169621786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,5120,0.023424888650576275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,6144,0.019160888261265226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,7168,0.0462773342927297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,4096,0.019346666004922655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,6144,0.04035022192531162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,3584,0.017107556263605755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,4096,0.014867555763986377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,5120,0.034745779302385114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,3072,0.015243555108706156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,4096,0.029395557112163965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,3584,0.013975999421543546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,3584,0.0262044436401791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,2560,0.013163555827405719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,3072,0.013088000317414602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,2048,0.011376000112957425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,3072,0.023539554741647508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,2560,0.012050666742854647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,1536,0.00907555553648207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,2560,0.02030311193731096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,2048,0.010851555400424533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,1024,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,2048,0.01759733259677887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,1536,0.00998488896422916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,768,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,1024,0.009327111144860586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,1024,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,512,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,768,0.008946666287051307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,768,0.010055111514197456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,256,0.004233777936961916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,5120,0.017056889004177518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,512,0.00831911133395301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,128,0.0042168887125121225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,512,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,2048,1536,0.016882666283183627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,64,0.0037786666717794207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,256,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,128,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1536,32,0.003994666867785984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,128,0.007346666521496243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,64,0.007350222104125553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,65536,0.21213244067298043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,32,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,1536,0.014486221803559197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,51200,0.14020800590515137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,65536,0.11168711715274388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,16384,0.05097155438529121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,65536,0.30000445577833385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,51200,0.0898524456553989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,12288,0.037645333343082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,16384,0.03074044320318434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,51200,0.24039644665188262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,16384,0.07937955856323242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,10240,0.03135022189882066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,256,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,12288,0.024125332633654278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,12288,0.05917244487338596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,10240,0.021920889616012573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,7168,0.02219999995496538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,10240,0.05052977800369263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,8192,0.01722666621208191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,6144,0.01966399947802226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,7168,0.01587466730011834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,5120,0.01665244499842326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,7168,0.03710577885309855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,6144,0.014452444182501899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1536,10240,0.06346311171849568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,6144,0.032807111740112305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,5120,0.013092444174819522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,3584,0.012488889197508493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,5120,0.02833333280351427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,4096,0.011720889144473605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,3072,0.010785777535703448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,4096,0.023384888966878254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,3584,0.011071999867757162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,2560,0.009581333233250512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,3072,0.0110880004035102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1536,65536,0.15432622697618273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,3072,0.018992889258596633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,2048,0.008200888832410177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,2560,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,1536,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,2048,0.009390222529570261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,2048,0.014523555835088095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,1024,0.005396444350481033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,8192,0.04161244299676683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,1536,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,1536,0.012096888489193387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,1024,0.008268444074524773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,1024,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,4096,0.013749333719412485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,512,0.0038480001191298165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,768,0.008936888641781276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,256,0.0032186667538351486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,512,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,512,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,128,0.003244444520937072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,256,0.007268444531493717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,256,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,64,0.0031786666562159858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,128,0.007094221810499827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,8192,0.024030221833123103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,32,0.00328533330725299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,128,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,2560,0.017186666528383892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,65536,0.14686044057210287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,64,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,32,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,51200,0.11318933963775635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,1024,768,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,16384,0.042062222957611084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,51200,0.08872799740897284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,65536,0.2616106669108073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,12288,0.031074666314654883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,16384,0.030593779351976182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,1024,768,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,10240,0.024919999970330134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,16384,0.06815110974841647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,12288,0.024481778343518574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,8192,0.019524445136388142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,12288,0.051628443929884166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,10240,0.02365422248840332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,7168,0.017638223038779367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,8192,0.018625777628686693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,1024,3584,0.021172444025675457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,6144,0.015712888704405892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,8192,0.03616799910863241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,7168,0.01587022178702884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,5120,0.013370667066839008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,6144,0.014652444256676568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,4096,0.011384000380833944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,6144,0.0284826656182607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,5120,0.01311377767059538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,3584,0.01016444464524587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,5120,0.024672889047198828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,4096,0.011981333295504252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,4096,0.020664888951513503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,3072,0.009098666409651438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,3584,0.01130488928821352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,3584,0.018927999668651156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,2560,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,3072,0.010815111299355825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,51200,0.20611466301812065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,2048,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,2560,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,2560,0.014930667148696052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,1536,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,10240,0.043859554661644824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,2048,0.009586667021115622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,2048,0.013088000317414602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,1024,0.005094222310516569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,1536,0.008771555291281806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,1536,0.011039999624093374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,768,0.004519111166397731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,7168,0.03251466817326016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,512,0.0038195554580953387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,1024,0.00904177791542477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,768,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,256,0.003169777699642711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,768,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,512,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,512,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,65536,0.11323732799953884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,256,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,256,0.006331555545330048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,64,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,128,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,32,0.0032008888406885993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,128,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,3072,0.017071111334694754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,65536,0.11615644560919867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,64,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,768,32,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,51200,0.09145510858959621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,16384,0.03825155562824673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,65536,0.10530755254957412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,51200,0.0852844450208876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,65536,0.21831022368537056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,12288,0.027671999401516382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,16384,0.028348445892333984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,768,1024,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,10240,0.02295111119747162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,16384,0.057077334986792676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,8192,0.01901511185699039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,12288,0.022621333599090576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,10240,0.020578665865792167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,7168,0.01701955497264862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,12288,0.04395111070738899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,8192,0.016175111134847004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,6144,0.015029332704014249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,10240,0.037552889850404524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,768,128,0.0031555555760860443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,8192,0.03108888864517212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,6144,0.013714666995737286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,7168,0.027968890137142603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,4096,0.010778666370444827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,6144,0.025088888075616624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,3584,0.009864889085292816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,5120,0.021373333202468023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,4096,0.011081777513027191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,3072,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,4096,0.01791288952032725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,3584,0.010464000205198923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,3584,0.016692444682121277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,2560,0.007891555627187094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,3072,0.010074666804737514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,3072,0.01483022173245748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,2048,0.006893333461549547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,2560,0.009650666680600908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,1536,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,51200,0.1710560056898329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,2560,0.013395555317401886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,1024,0.004538666870858935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,2048,0.00904888908068339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,2048,0.011717333561844297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,1536,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,768,0.003914666672547658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,1024,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,1536,0.009929777847396003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,512,0.0034764446318149567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,1024,0.008652444514963362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,768,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,7168,0.014816888504558139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,256,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,128,0.002559111143151919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,768,0.007382222347789341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,512,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,512,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,5120,0.012440889245933957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,32,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,128,0.007034666836261749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,256,0.005967111223273807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,128,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,65536,0.06396266486909655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,64,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,512,32,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,65536,0.10522933138741387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,65536,0.17414844036102295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,16384,0.021917333205540974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,51200,0.13621421655019125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,51200,0.0845866666899787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,12288,0.01793244481086731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,16384,0.027334221535258826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,16384,0.04651111033227709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,10240,0.01571555601225959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,5120,0.012816888590653738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,12288,0.02200444373819563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,8192,0.012871999707486896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,12288,0.0361697773138682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,512,64,0.002467555511328909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,512,256,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,7168,0.01165600038237042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,8192,0.01590755581855774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,10240,0.030932443009482488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,6144,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,8192,0.025410667061805725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,7168,0.014732443624072604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,5120,0.008640888664457533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,51200,0.05155822303560045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,7168,0.02277511192692651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,4096,0.007328000333574083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,6144,0.02050133380624983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,3584,0.006685333533419504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,4096,0.011072888970375061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,5120,0.017958222164048087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,3072,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,4096,0.015289778510729471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,3584,0.01038755559258991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,2560,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,3584,0.01421688828203413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,3072,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,2048,0.004574222283230888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,3072,0.012888888518015543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,2560,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,1536,0.004028444488843282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,2560,0.011600000162919363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,2048,0.010327999790509542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,1024,0.003220444545149803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,10240,0.020554666717847187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,1536,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,768,0.0031742221779293488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,1024,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,1024,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,6144,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,768,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,768,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,256,0.002558222247494592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,512,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,128,0.0022675556441148124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,512,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,256,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,64,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,128,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,32,0.002230222233467632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,128,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,65536,0.05161955621507433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,64,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,2048,0.009016888837019602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,32,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,51200,0.04090222385194566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,65536,0.10497688584857517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,256,1536,0.008984000318580205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,16384,0.015159999330838522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,65536,0.17360711097717285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,51200,0.08411733309427898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,12288,0.01275022245115704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,256,512,0.0028168888141711554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,51200,0.13574755191802979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,16384,0.02661955522166358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,10240,0.011156444748242697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,8192,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,12288,0.022103110949198406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,5120,0.012216889195972018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,7168,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,10240,0.02034222251839108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,12288,0.03539377782079909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,256,256,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,8192,0.015916445189052157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,7168,0.014720888601409065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,8192,0.0248364441924625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,5120,0.008512000242869059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,6144,0.013463111387358772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,7168,0.022682666778564453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,4096,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,6144,0.020201777418454487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,5120,0.012113778127564324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,3584,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,5120,0.01756000022093455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,3072,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,3584,0.010440888504187265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,3584,0.014127110441525778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,2560,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,3072,0.010055999788973067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,3072,0.01276533305644989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,2048,0.004547555413511065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,2560,0.00959288908375634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,2560,0.011462222370836469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,16384,0.04594488938649496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,1536,0.0038906666967603895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,2048,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,2048,0.009004444711738164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,6144,0.00998933364947637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,1024,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,1536,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,768,0.0029004443850782183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,1024,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,1024,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,512,0.0025395556456512874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,4096,0.011063111325105032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,768,0.00693866651919153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,4096,0.015237333046065437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,256,0.00221422231859631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,512,0.007326222128338284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,128,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,256,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,256,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,64,0.001981333312061098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,128,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,128,32,0.0022044444663657085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,64,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,10240,0.030064000023735896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,65536,0.0474586652384864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,32,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,51200,0.03769955701298184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,1536,0.00906399968597624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,16384,0.012091555529170565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,12288,0.012780444489585029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,768,0.007592889169851939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,65536,0.17266310585869682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,51200,0.13559199704064265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,16384,0.04583377639452616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,12288,0.034847110509872437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,7168,0.009604444106419882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,128,512,0.005966222120655908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,6144,0.009207111265924241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,10240,0.02977777851952447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,8192,0.024843555357721116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2049,128,128,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,4096,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,6144,0.02011999984582265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,3584,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,5120,0.01774222155412038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,3072,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,2560,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,4096,0.015198222464985318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,2048,0.004567111117972268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,3584,0.014150222142537435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,3072,0.012726222475369772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,1536,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,2560,0.011477332976129321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,2048,0.010075555907355415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,10240,0.011420444481902652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,768,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,512,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,8192,0.010407999985747868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,1536,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,256,0.0021679999513758556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,1024,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,128,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,64,0.0018648888087934917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,768,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,32,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,512,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,5120,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,256,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,65536,0.04921155505710178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,128,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,51200,0.037244445747799344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,7168,0.022621333599090576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,64,0.005030222237110138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,12288,0.008616888688670265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,10240,0.007954667011896769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,64,32,0.00519466648499171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,8192,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,16384,0.045913779073291354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,51200,0.13534311453501383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,12288,0.03528177738189697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,7168,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,6144,0.006319111006127463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,10240,0.029836446046829224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,5120,0.010870222416188983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,8192,0.024792888098292883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,4096,0.009126221968068017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,3584,0.008105778031879002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,64,1024,0.0031804444475306403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,3072,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,7168,0.022396443618668452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,6144,0.020107555720541213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,2560,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,5120,0.017641777793566387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,2048,0.004274666723277834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,4096,0.015180443723996481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,1536,0.0037786666717794207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,1024,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,3584,0.014160000615649752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,768,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,3072,0.01276000009642707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,2560,0.011543110840850405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,512,0.0024702221983008915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,2048,0.010061333576838175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,1536,0.008854221966531541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,256,0.0022044444663657085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,1024,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,128,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,64,0.0018604445374674266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,16384,0.010081777969996134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,768,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2049,32,32,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,512,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,256,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,128,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,64,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,65536,0.1717848910225762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2049,32,32,0.005025777965784073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,16384,2.575807147555881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,16384,2.914907455444336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,16384,0.7873519791497124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,51200,2.996119181315104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,16384,2.403775956895616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,12288,1.833298577202691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,12288,2.050567203097873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,12288,0.7212702433268229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,10240,1.5734355714586046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,10240,0.5069022178649902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,10240,1.6553671095106337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,12288,1.7939457363552516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,8192,1.3107439676920574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,8192,0.4125520123375787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,51200,7.8625132242838545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,10240,1.5139928393893773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,8192,1.4305199517144098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,8192,1.1907004250420463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,7168,1.1900640063815646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,7168,1.201006253560384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,6144,1.0208524068196614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,6144,0.3222240077124702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,7168,1.0149866739908855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,6144,1.0063582526312935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,5120,0.8261626561482748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,7168,0.3682862122853597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,6144,0.8738462130228678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,5120,0.27566756142510307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,5120,0.828670236799452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,4096,0.651076422797309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,4096,0.6424178017510308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,4096,0.22780354817708334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,3584,0.47784799999660915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,5120,0.7084462377760145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,3584,0.20385066668192545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,51200,11.696917215983072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,4096,0.5663093460930718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,3584,0.5889857610066732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,3072,0.45207468668619794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,3584,0.5173893504672581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,3072,0.1832533280054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,3072,0.5049973593817817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,2560,0.4005555576748318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,3072,0.44677421781751847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,2560,0.41525154643588597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,2048,0.3241644435458713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,2048,0.13418400287628174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,2048,0.33405600653754336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,2560,0.1601946618821886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,2560,0.39717332522074383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,2048,0.32696710692511666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,1536,0.28743823369344074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,1024,0.16835378275977242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,1536,0.10908266570832993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,768,0.13319110870361328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,1024,0.08755022287368774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,1536,0.2461946540408664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,1024,0.2016622225443522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,1024,0.18808266851637098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,512,0.09596088859770034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,51200,8.179688347710504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,768,0.07691466808319092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,512,0.06976799832450019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,768,0.19227288828955758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,256,0.07531555493672688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,512,0.1381831169128418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,128,0.06998844279183282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,512,0.12605155838860407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,256,0.12390044000413682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,256,0.06400889158248901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,1536,0.23798934618631998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,128,0.05920000208748711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,128,0.11592266294691299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,256,0.09862133529451157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,128,0.07521422041787042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,64,0.1187351147333781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,32,0.11774310800764297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,768,0.1603040032916599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,64,0.048002666897243924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,65536,32,0.0492835541566213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,65536,2.943481869167752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,51200,2.2330854203965926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,51200,5.955334981282552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,16384,2.398308436075846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,65536,7.8035634358723955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,16384,0.6112586657206217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,16384,2.2391014099121094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,12288,1.5561066733466253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,51200,9.372439914279514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,12288,1.7373920016818578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,16384,1.889802720811632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,12288,0.47499646080864805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,10240,1.328456030951606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,10240,0.39715109931098086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,65536,11.432426452636719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,12288,1.406038178337945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,51200,6.568562825520833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,8192,0.31748888227674693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,8192,1.0035520129733615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,8192,1.1503911548190646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,10240,1.1648257573445637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,10240,1.3415181901719835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,7168,0.9077235327826606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,7168,0.2910737726423475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,8192,0.9232186211480035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,7168,0.928577740987142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,6144,0.7708906597561307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,6144,0.25464267200893825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,65536,8.271241082085504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,6144,0.753753768073188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,7168,0.7956524425082736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,5120,0.6375350952148438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,5120,0.2227395640479194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,5120,0.6301715638902452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,6144,0.676270220014784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,4096,0.5270657539367676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,4096,0.20339288976457384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,3584,0.4632862408955892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,4096,0.5012586911519369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,3584,0.1743182208802965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,5120,0.5661048889160156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,3584,0.4313991069793701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,4096,0.46150933371649844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,3072,0.14309688409169516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,2560,0.30370667245652944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,3072,0.4724719789293077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,3072,0.3557164404127333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,3584,0.41163910759819883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,2560,0.3209475676218669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,2560,0.12629332807328966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,2048,0.27069422933790416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,1536,0.16490844885508218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,2560,0.2971031135982937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,2048,0.10918844408459133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,2048,0.2797706656985813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,1024,0.1314106649822659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,1536,0.08526488807466294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,3072,0.362519105275472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,1024,0.06879822413126628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,1024,0.15912622875637478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,1536,0.20943466822306314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,1536,0.20329777399698892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,1024,0.152728001276652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,768,0.11485955450269912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,512,0.07895200120078193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,768,0.061335113313463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,256,0.05785955323113335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,512,0.055247111452950366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,768,0.12674666775597468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,768,0.1234382258521186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,128,0.045372443066702954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,512,0.1101413302951389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,512,0.10041866699854533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,256,0.05046755406591627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,64,0.03830844495031569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,256,0.09697777695126003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,51200,128,0.048413332965638906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,51200,32,0.03912977708710565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,128,0.09187733464770848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,128,0.060886224110921226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,64,0.09277333153618707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,51200,32,0.09336444404390122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,2048,0.24949423472086588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,51200,256,0.079894224802653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,65536,0.8970684475368924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,51200,2.0525574154324002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,51200,0.6943377918667264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,16384,0.8602728843688965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,65536,2.655327055189345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,16384,0.22217599550882974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,51200,2.7615928649902344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,12288,0.4940444628397624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,16384,0.7130586836073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,12288,0.16306222809685603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,65536,3.776543935139974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,10240,0.3550559944576687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,16384,0.6094631089104546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,12288,0.5097724596659342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,51200,2.1160009172227645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,12288,0.44422043694390195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,10240,0.14341066943274602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,8192,0.3230737845102946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,8192,0.1137804455227322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,10240,0.42859554290771484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,65536,2.7152311537000866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,7168,0.2722168763478597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,8192,0.3374151123894586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,10240,0.36921244197421604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,8192,0.3230115572611491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,6144,0.22980801264444986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,7168,0.10270044538709854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,6144,0.08865511417388916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,6144,0.2503511110941569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,7168,0.299544890721639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,5120,0.18310311105516222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,7168,0.26298311021592885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,6144,0.22737956047058105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,4096,0.14703110853830972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,5120,0.07589955462349786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,5120,0.2188168896569146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,4096,0.06350933180914985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,3584,0.12568267186482748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,4096,0.17279199759165445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,5120,0.19221866130828857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,4096,0.16097422440846762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,3584,0.05678844451904297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,3584,0.1466346714231703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,3584,0.1429493294821845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,2560,0.11205155319637722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,3072,0.05120355553097195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,3072,0.12937245104047987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,2048,0.08026311132642958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,3072,0.1235155529446072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,2560,0.04483733243412442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,2560,0.11112977398766412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,1536,0.06136711438496908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,2048,0.03919555412398444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,2560,0.10888889100816514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,2048,0.09210577938291763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,1024,0.04468533396720886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,1536,0.03229422370592753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,1536,0.07356977462768555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,768,0.031926221317715116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,1536,0.07421333260006376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,1024,0.0266275554895401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,1024,0.056001775794559054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,512,0.02321333355373806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,1024,0.05901688999599881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,768,0.02418311105834113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,3072,0.10735466745164658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,768,0.04499377806981405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,256,0.01772711177666982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,768,0.0490515563223097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,512,0.02237777743074629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,128,0.016311999824311998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,512,0.03723288906945123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,256,0.019920888874265883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,512,0.04162755608558655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,64,0.013658666776286231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,256,0.033407111962636314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,16384,32,0.014105778601434497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,128,0.019298666053348117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,256,0.033384889364242554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,128,0.03142755561404758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,128,0.025785777303907607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,64,0.031954665978749595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,32,0.03180977702140808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,2048,0.09100266959932114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,65536,0.7078791194491916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,51200,0.5689999792310927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,51200,1.568384912278917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,16384,0.6011066436767578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,65536,2.1268533070882163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,16384,0.5575591193305122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,16384,0.1873262193467882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,51200,2.289505852593316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,12288,0.3894764317406549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,16384,0.46175824271308047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,65536,2.637441847059462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,12288,0.3897262149386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,12288,0.13205599784851074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,10240,0.33764534526401097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,51200,1.603949334886339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,12288,0.346085336473253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,10240,0.11026133431328668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,8192,0.27419643931918675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,10240,0.3323804537455241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,8192,0.08812444739871556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,10240,0.2984986570146349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,65536,2.0924898783365884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,8192,0.35438044865926105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,7168,0.07904533545176189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,6144,0.17174932691786024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,8192,0.26080799102783203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,7168,0.23880444632636177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,5120,0.14217155509524876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,7168,0.21007911364237467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,6144,0.06953510973188612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,6144,0.21181066830952963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,4096,0.13329066170586482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,5120,0.05886222256554497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,5120,0.16736355092790392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,5120,0.15333155790964761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,3584,0.10539288653267755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,4096,0.04996622271007962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,4096,0.1277786625756158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,3584,0.0448293321662479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,3072,0.09942133559121026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,3584,0.11477243900299072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,7168,0.20744889312320283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,3584,0.11396000120374893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,2560,0.08458933565351699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,3072,0.04007644454638163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,3072,0.10024888647927178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,3072,0.10099289152357314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,2048,0.06273066997528076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,2560,0.035806222094429865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,2560,0.08640178044637044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,1536,0.04745955599678887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,2048,0.0304097765021854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,2048,0.07114311059315999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,6144,0.18419022030300566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,2048,0.07436355617311266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,1024,0.03793333305252923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,4096,0.12900888919830322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,1536,0.05688977903789944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,1536,0.025174222058720056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,768,0.027648000253571406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,1536,0.06150044335259331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,1024,0.021009777983029682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,1024,0.043452445003721446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,512,0.02071822186311086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,768,0.019023999571800232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,1024,0.048115554783079356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,256,0.01536711057027181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,768,0.03547022077772353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,768,0.04150844282574124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,512,0.01725955473052131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,512,0.02884977724817064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,128,0.01221777747074763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,512,0.03447644578086005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,256,0.016008888681729633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,256,0.02570666703912947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,64,0.011053333679835001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,256,0.02747111188040839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,12288,32,0.011403555671374003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,128,0.01585422290696038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,128,0.024634665913052026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,128,0.02128177715672387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,2560,0.08675644132826064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,64,0.02457511093881395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,32,0.02444177700413598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,65536,0.5708950890435113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,51200,1.2096444235907662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,65536,1.547284444173177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,51200,0.47935909695095485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,16384,0.4687955644395616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,16384,0.1490559975306193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,12288,0.3619839880201552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,16384,0.4634382459852431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,51200,1.9095191955566406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,65536,2.0545306735568576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,16384,0.382842673195733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,12288,0.11024977763493855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,12288,0.33027733696831596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,51200,1.340651512145996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,10240,0.09385777844323052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,8192,0.22927021980285645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,12288,0.41118489371405703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,65536,1.6386595831976996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,10240,0.2503253353966607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,8192,0.07647022273805407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,8192,0.25726400481330025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,6144,0.17762488789028594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,7168,0.06852177778879802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,8192,0.1986924409866333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,7168,0.23525688383314344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,7168,0.17640444967481825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,5120,0.13945689466264513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,6144,0.05980177720387777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,6144,0.17296444045172799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,6144,0.15370933214823404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,10240,0.2795235580868191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,5120,0.05135200089878506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,4096,0.10584800110922919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,5120,0.1357440021302965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,3584,0.09176266855663723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,10240,0.27339198854234487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,5120,0.1364835500717163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,4096,0.10938666926489936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,3072,0.0817777779367235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,3584,0.03728266557057699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,4096,0.10971911085976495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,7168,0.20403377215067545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,3584,0.09653244415918986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,2560,0.06932710938983493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,3584,0.09967733091778225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,3072,0.08524977498584324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,2048,0.05535911189185249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,2560,0.02987288766437107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,2560,0.07397066884570651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,2560,0.0763911141289605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,1536,0.04035200013054742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,2048,0.02587466604179806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,2048,0.06040177742640177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,2048,0.06537422206666735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,1024,0.02971200148264567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,1536,0.04907733201980591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,1536,0.021398221453030903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,1536,0.052575111389160156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,768,0.02241955531968011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,1024,0.018028444714016385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,4096,0.042875554826524526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,1024,0.03717511230044895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,512,0.016400888562202454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,1024,0.042428443829218544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,768,0.016526222229003906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,768,0.030716445710923936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,256,0.012501333322789935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,768,0.036385777923795916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,512,0.015267555912335714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,3072,0.03346755438380771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,512,0.024731554918819006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,128,0.009986666341622671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,3072,0.08790133396784465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,512,0.03044355577892727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,256,0.014132445057233175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,64,0.008970666262838576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,10240,32,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,128,0.014114666316244336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,128,0.021264889174037512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,128,0.01850400037235684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,64,0.021355556117163763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,32,0.021301334102948506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,65536,0.49330843819512266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,256,0.02199733257293701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,256,0.024323556158277724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,65536,1.3378933800591364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,51200,1.2157298194037545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,16384,0.3151013321346707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,51200,0.3929146660698785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,16384,0.1180151171154446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,51200,1.3967599868774414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,12288,0.3022079997592502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,65536,1.397159152560764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,12288,0.10567822058995564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,12288,0.2623608907063802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,12288,0.2505884435441759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,10240,0.23349422878689238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,51200,1.0540302064683702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,10240,0.07567466629876031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,8192,0.1758577823638916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,10240,0.27010220951504177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,10240,0.20086044735378691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,8192,0.06757866673999362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,7168,0.15922666920555964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,65536,1.9507901933458116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,16384,0.4117573367224799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,8192,0.16381066375308567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,7168,0.056065777937571205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,16384,0.34556089507208926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,7168,0.15296088324652776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,5120,0.1099555558628506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,6144,0.04923911227120293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,6144,0.13372800085279676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,7168,0.15012178156110975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,4096,0.07615821891360812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,6144,0.12790666686164007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,5120,0.1114026705423991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,5120,0.04189777705404493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,3584,0.07769155502319336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,4096,0.035954667462242976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,5120,0.10997955666648017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,4096,0.09075378047095405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,3072,0.0668577750523885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,3584,0.031172444423039753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,3584,0.08054222000969781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,2560,0.05830044216579861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,3584,0.079821334944831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,3072,0.02810133496920268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,3072,0.07015377945370145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,8192,0.20741421646542021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,2048,0.04042311178313361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,3072,0.07374933030870226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,2560,0.02572444412443373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,2560,0.06096000141567654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,1536,0.03334844443533156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,2048,0.022275555464956496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,2560,0.06288799974653456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,2048,0.05012977785534329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,2048,0.053901334603627525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,1536,0.019038221902317472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,1536,0.04029511080847846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,768,0.018738667170206707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,1536,0.04551733202404446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,1024,0.01583377851380242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,1024,0.03128266665670607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,512,0.013487111363146039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,1024,0.035798221826553345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,768,0.014554666148291694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,768,0.02590311070283254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,256,0.010727110836240979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,4096,0.08976355526182388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,768,0.03083733386463589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,512,0.013448889056841532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,512,0.02050488856103685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,128,0.009187555147541894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,512,0.02533955540921953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,256,0.012683555483818054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,256,0.01868799991077847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,64,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,256,0.021394666698243883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,128,0.01257777793539895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,6144,0.12975554996066624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,128,0.017867555220921833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,128,0.016919111212094624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,1024,0.023727999793158636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,64,0.018238221605618794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,32,0.017887110511461895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,65536,0.4483742184109158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,51200,0.9620115492078992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,8192,32,0.008022222254011366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,65536,1.3314844767252605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,16384,0.30735821194118923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,51200,1.257537735833062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,51200,0.3457831011878119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,16384,0.3252204524146186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,16384,0.1128951046201918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,12288,0.2559102111392551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,65536,1.3175724877251518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,16384,0.3040257824791802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,12288,0.2296968830956353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,12288,0.08567021952735053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,10240,0.21162045001983643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,12288,0.23562844594319662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,10240,0.06900622447331746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,51200,1.0125190946790907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,8192,0.1752586629655626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,10240,0.20758488443162706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,10240,0.19436089197794595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,8192,0.056073778205447726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,8192,0.16170844766828749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,65536,1.5194364123874242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,7168,0.04973333411746555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,8192,0.16087199581993952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,6144,0.12861422697703043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,7168,0.1406666702694363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,7168,0.146415114402771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,5120,0.1046302252345615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,6144,0.044017778502570264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,6144,0.12315200434790717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,4096,0.08410133255852593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,6144,0.12352889113956028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,5120,0.037843555212020874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,5120,0.09989511304431492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,3584,0.06558310985565186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,5120,0.10575111044777764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,4096,0.03228177626927694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,4096,0.08127911223305596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,3072,0.06126044193903605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,3584,0.029363556040657893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,3584,0.07236444287829928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,2560,0.048044443130493164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,3072,0.02716355522473653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,3584,0.07864977916081746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,3072,0.06411466995875041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,2048,0.03748355640305413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,2560,0.024381332927280005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,3072,0.07111289103825887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,1536,0.02907022171550327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,7168,0.15442132949829102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,2048,0.020794666475719877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,2560,0.06156355804867215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,1024,0.020958221620983545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,1536,0.017722666263580322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,1536,0.043325334787368774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,1024,0.028646220763524372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,768,0.016122667325867545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,1024,0.01482488877243466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,1024,0.034648001194000244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,4096,0.08812710973951553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,768,0.013879111243618859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,512,0.013316444224781461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,768,0.02950577934583028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,512,0.013006221916940479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,512,0.019208000765906442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,2560,0.05520800087187025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,256,0.00963644435008367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,512,0.024673778149816725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,128,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,2048,0.04605866803063286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,256,0.012087110843923358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,2048,0.05224711034033033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,64,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,256,0.01719377769364251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,1536,0.037187556425730385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,7168,32,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,256,0.01977777812216017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,128,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,128,0.016286222471131217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,128,0.016474665866957772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,64,0.016513778103722464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,32,0.016172443826993305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,768,0.0236835562520557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,65536,0.3859235445658366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,65536,1.0423155890570748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,51200,0.8697928852505155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,51200,0.301800012588501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,16384,0.2672097682952881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,65536,1.2379635704888237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,51200,0.9880906211005317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,16384,0.351781341764662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,16384,0.10068977541393703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,12288,0.19091910786098906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,16384,0.27749066882663304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,65536,1.1439796023898654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,12288,0.21211555269029406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,51200,0.8624666531880697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,12288,0.21488444010416666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,8192,0.13233155674404568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,10240,0.17120088471306694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,10240,0.19946489069196913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,8192,0.05013066530227661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,8192,0.13675644662645128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,7168,0.11421067184872097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,8192,0.15166132979922825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,6144,0.09411378039254083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,7168,0.04520533482233683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,7168,0.12034489048851861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,7168,0.1268035570780436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,6144,0.039035555389192365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,5120,0.07985955476760864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,12288,0.07973955737219916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,6144,0.1077466673321194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,4096,0.06733866532643636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,6144,0.11088444126976861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,5120,0.03439911206563314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,5120,0.08858666817347209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,3584,0.060524443785349526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,10240,0.061894224749671094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,4096,0.029168887270821467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,5120,0.09433511230680679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,3072,0.055088000165091626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,4096,0.07181511322657268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,3584,0.026867555247412786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,4096,0.07888088623682658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,2560,0.04043022129270766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,3584,0.06373955806096394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,3072,0.02439733346303304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,3584,0.07125244537989299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,2048,0.03261333372857835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,3072,0.05566755268308851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,2560,0.0222053329149882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,3072,0.06328533093134563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,2560,0.04860088891453213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,1536,0.025968889395395916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,2048,0.019176888797018263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,2560,0.05547555618815952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,1024,0.018615111708641052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,2048,0.04071111149258084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,1536,0.016720889343155753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,2048,0.04706755611631605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,768,0.015776889191733468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,1536,0.03316088848643833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,1024,0.014157333307796054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,10240,0.1640711095598009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,512,0.011628444823953839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,1024,0.025095999240875244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,1024,0.031222220924165513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,768,0.013083555632167392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,768,0.02117777698569828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,256,0.010991999672518836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,512,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,768,0.026143999563323125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,128,0.009378666679064432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,256,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,512,0.022479999396536086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,64,0.008754666480753157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,256,0.015304888288180033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,6144,32,0.008951110972298516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,128,0.011431111229790581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,128,0.014850666125615438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,128,0.015211555692884656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,64,0.014789332946141561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,32,0.014774221513006421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,1536,0.03930044505331252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,65536,0.3313431209988064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,51200,0.6272355715433756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,512,0.01696266730626424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,65536,1.0383271111382377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,256,0.01925155520439148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,16384,0.21872356202867296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,65536,1.1926551394992404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,51200,0.27091556125217015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,12288,0.1762453317642212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,16384,0.08044089211357965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,16384,0.33237067858378094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,65536,0.9533280266655816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,51200,0.8331066767374674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,10240,0.1396053367190891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,12288,0.06305599874920316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,16384,0.23549066649542916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,12288,0.18764711750878227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,8192,0.11667022440168594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,51200,0.7293253474765353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,12288,0.18756533993615043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,10240,0.15087466769748265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,10240,0.15684621863894993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,8192,0.044341332382626004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,8192,0.1204648945066664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,8192,0.1256506707933214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,6144,0.08864711390601264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,7168,0.04036088784535726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,7168,0.10505155722300212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,5120,0.06892800331115723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,7168,0.11223733425140381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,6144,0.08922488821877374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,6144,0.0966871115896437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,5120,0.07631466786066692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,4096,0.05630044142405192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,5120,0.029917332861158583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,5120,0.0834524432818095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,3584,0.04792177677154541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,4096,0.025621333056026038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,10240,0.058902223904927574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,4096,0.062393777900271945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,4096,0.06933066580030653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,7168,0.09379555781682332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,3584,0.02367111047108968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,3584,0.055776887469821505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,2560,0.03220977716975742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,3584,0.062409778436024986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,3072,0.021673777037196692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,3072,0.04903911219702827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,2048,0.026245334082179602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,2560,0.019528888993793063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,6144,0.035100443495644465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,2560,0.0428746673795912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,1536,0.02016177773475647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,2560,0.04889333248138428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,2048,0.016945777667893302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,2048,0.03566044569015503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,1024,0.0151582227812873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,2048,0.04184177849027845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,1536,0.014828445182906257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,768,0.012301333248615265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,1536,0.03508088986078898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,1024,0.012890666723251343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,1024,0.02200177808602651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,512,0.009305777649084726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,3072,0.03913066784540812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,1024,0.027310222387313843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,768,0.01204622205760744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,768,0.01885333326127794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,256,0.0068862222962909276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,768,0.023579556081030104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,512,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,512,0.015136889285511441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,128,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,256,0.010564444793595208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,256,0.013580444786283704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,512,0.02038666605949402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,64,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,256,0.017880888448821176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,5120,32,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,3072,0.05656355619430542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,128,0.01051644484202067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,128,0.013200889031092325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,128,0.014223999447292753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,64,0.013144888811641268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,32,0.013066666821638743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,1536,0.029189334975348577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,65536,0.2956799930996365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,65536,0.6164799796210395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,51200,0.5438542366027832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,16384,0.17479911115434435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,51200,0.22921421792772081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,12288,0.1307351059383816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,16384,0.07043466965357463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,65536,0.9414595497979058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,65536,0.7823440233866373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,12288,0.06086044179068672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,16384,0.2535697884029812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,12288,0.14764532777998182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,10240,0.11473066276974148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,12288,0.1560764445198907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,8192,0.09020444419648911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,10240,0.047136889563666455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,10240,0.13239288330078125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,10240,0.13417777750227186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,8192,0.03715288970205519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,8192,0.09825955496893989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,8192,0.11304622226291233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,7168,0.033257779147889875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,7168,0.08745244476530288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,7168,0.09517955780029297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,16384,0.20322044690450033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,5120,0.052447110414505005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,6144,0.029631111356947158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,6144,0.07483733362621732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,51200,0.6134737862481011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,6144,0.08315733406278822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,4096,0.0480275551478068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,5120,0.026219555073314246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,5120,0.06351821952395968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,5120,0.0720391141043769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,3584,0.03966666592491998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,4096,0.0225582222143809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,4096,0.059803558720482715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,7168,0.0827724470032586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,3072,0.03385688861211141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,3584,0.020592888196309406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,3584,0.046726223495271474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,3584,0.05537777807977465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,6144,0.06765333149168226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,2560,0.02899466620551215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,3072,0.018765333626005385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,51200,0.675797356499566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,2048,0.023698666029506262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,3072,0.048957334624396436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,2560,0.01703555550840166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,2560,0.036045332749684654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,1536,0.018565333551830716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,2560,0.04291022155019972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,2048,0.01521511044767168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,2048,0.030228445927302044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,1536,0.013600889179441663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,2048,0.03707199957635667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,1536,0.02414488957987891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,1024,0.014469333820872836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,768,0.011607999602953592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,4096,0.05224355392985874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,1024,0.011776000261306763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,1024,0.019137778215938143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,512,0.008650666309727563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,1024,0.023822221491071913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,768,0.010963555839326648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,768,0.021306667062971327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,512,0.010272888673676385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,256,0.006341333190600078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,3072,0.04150044586923387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,512,0.013325333595275879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,512,0.018224888377719455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,128,0.005981333139869902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,256,0.009663110805882348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,256,0.012042666474978128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,256,0.016533333394262526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,32,0.0063324446479479475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,128,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,128,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,128,0.013391999734772576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,64,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,32,0.011401777466138205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,1536,0.030881779061423406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,768,0.016191111670600045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,65536,0.2670604387919108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,65536,0.6496355798509386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,51200,0.45128711064656574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,16384,0.19149333900875518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,65536,0.771033763885498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,51200,0.2115093337164985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,4096,64,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,51200,0.605695989396837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,12288,0.12280978096856011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,16384,0.17171822653876412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,65536,0.7771484586927625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,10240,0.09672977526982625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,12288,0.05264977614084879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,16384,0.22084532843695748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,12288,0.12910844220055476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,51200,0.6021751297844781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,8192,0.08663911289638943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,12288,0.15758132934570312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,10240,0.11510754956139459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,7168,0.07338577508926392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,10240,0.12960888279808894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,8192,0.0890551143222385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,6144,0.06213688850402832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,8192,0.1074942217932807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,7168,0.07958133353127374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,7168,0.09474488761689927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,6144,0.028675556182861328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,5120,0.05246666736072964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,6144,0.0685359968079461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,6144,0.08335110876295303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,16384,0.06544088655047946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,4096,0.04271466533342997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,5120,0.024900444679790076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,5120,0.0581644442346361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,3584,0.03805333375930786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,5120,0.07062844435373943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,4096,0.021382222572962444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,4096,0.04818400078349643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,10240,0.04460533459981283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,3072,0.03260711166593764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,4096,0.059875554508633085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,3584,0.019759999381171334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,3584,0.04373244444529215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,8192,0.0362462235821618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,3072,0.01847555571132236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,3072,0.03843199875619676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,7168,0.031952887773513794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,2048,0.023939554889996845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,2560,0.016566221912701923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,2560,0.03408088948991563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,2560,0.041800889703962535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,2048,0.014848000473446317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,2048,0.02790400054719713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,2048,0.03584266702334086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,1536,0.023062222533755835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,1536,0.013370667066839008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,1024,0.014489778214030795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,1536,0.02930844492382473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,1024,0.011411555939250521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,1024,0.017664000391960144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,768,0.011232888533009423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,2560,0.028753777345021565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,1024,0.023353776997990076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,512,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,3584,0.053145776192347206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,768,0.01073155552148819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,768,0.015136000182893542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,3072,0.04779555400212606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,256,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,768,0.020777778493033517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,512,0.0124195557501581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,512,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,128,0.00574311117331187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,512,0.01762488815519545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,1536,0.019110222657521565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,256,0.009591110878520543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,64,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,256,0.011119999819331698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3584,32,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,128,0.00960533320903778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,256,0.01538133290078905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,128,0.01096088853147295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,128,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,64,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,32,0.010662222074137794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,65536,0.21933155589633516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,51200,0.397416008843316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,65536,0.5396880043877496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,16384,0.1305653386645847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,51200,0.18287021583980986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,12288,0.10130221976174249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,16384,0.06381244129604764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,65536,0.6746515697903104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,16384,0.16148800320095485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,10240,0.09357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,51200,0.5034311082628038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,12288,0.04538044333457947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,12288,0.11386399798923069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,8192,0.07666133509741889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,10240,0.039493332306543984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,10240,0.10133244593938191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,10240,0.10884088940090603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,7168,0.0638399985101488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,8192,0.03104355600145128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,8192,0.0787573324309455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,8192,0.08869600296020508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,6144,0.05268889003329807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,7168,0.029895110262764826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,7168,0.07021333111657037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,7168,0.07961244715584649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,5120,0.044059554735819496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,51200,0.5272693104214138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,6144,0.026762665973769292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,6144,0.06944888830184937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,65536,0.6325564384460449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,16384,0.18734844525655112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,4096,0.037112000915739275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,5120,0.022670222653283015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,3584,0.0329039990901947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,5120,0.052988443109724254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,4096,0.019685332973798115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,12288,0.12911910480923122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,3072,0.027950220637851294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,4096,0.05047466688685947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,3584,0.017688888642523024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,3584,0.03892799880769517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,2560,0.02363555630048116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,3584,0.046331554651260376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,3072,0.01642933322323693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,3072,0.03464444478352865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,2048,0.01962577799956004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,2560,0.014565333724021912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,3072,0.04109422365824381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,2560,0.0299635562631819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,1536,0.01407199932469262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,2560,0.03650310966703627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,2048,0.013316444224781461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,2048,0.02532177832391527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,1024,0.010650667051474253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,6144,0.061271111170450844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,1536,0.012091555529170565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,1536,0.020467556185192533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,768,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,1536,0.026081777281231348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,1024,0.010388444695207808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,1024,0.016069332758585613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,512,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,5120,0.060192889637417264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,1024,0.020940444535679285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,768,0.009810667071077559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,768,0.013785777820481194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,256,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,768,0.018930666976504855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,512,0.009335111412737105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,512,0.011457777685589261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,512,0.01754666699303521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,128,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,256,0.008767999708652496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,256,0.01034311122364468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,64,0.004823110997676849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,256,0.01484266585773892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,128,0.009007111191749573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,3072,32,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,128,0.009750222166379293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,128,0.011977777712874942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,64,0.009683555199040307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,32,0.009888889061080085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,65536,0.4479955567253961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,65536,0.20886311266157362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,51200,0.35996799998813206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,4096,0.043571554952197604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,65536,0.5884479946560329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,16384,0.10819733142852783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,51200,0.1695528825124105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,65536,0.6272462209065756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,51200,0.4738782246907552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,16384,0.13307289282480875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,16384,0.05526577764087253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,16384,0.16446755992041692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,51200,0.4908880127800836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,10240,0.07282044490178426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,12288,0.04297777679231432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,12288,0.10309511423110962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,2048,0.03156088789304098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,8192,0.060871111022101514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,12288,0.12641333209143743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,10240,0.03668266534805298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,7168,0.05018577641910977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,10240,0.08462222417195638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,8192,0.029574222034878198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,8192,0.06891555256313749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,6144,0.04407111141416761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,7168,0.026625777284304302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,7168,0.062020447519090444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,8192,0.08837778038448757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,5120,0.03532444437344869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,7168,0.07797688908047147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,6144,0.023689778314696416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,4096,0.029475556479560003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,6144,0.053320888015958995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,6144,0.068686220380995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,5120,0.0460382236374749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,3584,0.023846222294701472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,5120,0.05908444192674425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,12288,0.08727644549475776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,4096,0.018207111292415194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,4096,0.03790844480196635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,3072,0.022261333134439256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,4096,0.04968533251020643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,3584,0.016768000192112394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,2560,0.01756977703836229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,3584,0.03414222266938951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,3072,0.015544000599119397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,3584,0.04488889045185513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,2048,0.014463999205165438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,3072,0.04048888882001241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,2560,0.01423911088042789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,2560,0.02649066680007511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,1536,0.011769777370823754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,2560,0.03580710954136319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,2048,0.013076444466908773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,2048,0.02237777743074629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,1024,0.009361777868535783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,2048,0.030068443881140813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,1536,0.011367999845080905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,1536,0.01829866733815935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,768,0.0074728892909155945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,1536,0.025274667474958632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,1024,0.010223999619483948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,5120,0.021187555458810594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,1024,0.014427555931939019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,512,0.006320888797442119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,1024,0.020503999458418954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,768,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,768,0.012768000364303589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,256,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,768,0.018532445033391316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,512,0.01609066625436147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,128,0.004566222015354368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,256,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,3072,0.030444443225860596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,256,0.01386844449573093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,64,0.004260444392760594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,128,0.008640888664457533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2560,32,0.004510222209824456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,10240,0.10658044285244411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,128,0.011047999891969891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,64,0.009201777478059133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,32,0.00908799966176351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,65536,0.34066398938496906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,65536,0.16353866789076063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,512,0.01036622209681405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,65536,0.4734675619337294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,256,0.009657777845859528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,51200,0.2697439988454183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,65536,0.48682488335503477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,16384,0.08734933535257976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,128,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,51200,0.13083822197384304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,51200,0.37858133845859104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,12288,0.06979999939600627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,16384,0.04344088832537333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,16384,0.11522489123874241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,51200,0.38370222515530056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,10240,0.05949155489603678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,16384,0.143321779039171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,12288,0.033590220742755465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,12288,0.08556711011462742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,8192,0.04786400000254313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,10240,0.030551999807357788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,12288,0.10009155670801799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,7168,0.040525333748923406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,10240,0.07289244731267293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,8192,0.023898666103680927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,10240,0.08582933081520928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,6144,0.03539289037386576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,512,0.009101333717505137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,8192,0.05897777610354953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,7168,0.021531555387708876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,5120,0.029007110330793593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,8192,0.07106488943099976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,7168,0.052701334158579506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,6144,0.019620445039537217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,4096,0.023738667368888855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,7168,0.06351200077268812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,5120,0.01731911136044396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,6144,0.05547644363509285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,3584,0.021687111920780603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,5120,0.03950488898489211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,4096,0.015406222807036506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,5120,0.04878133204248217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,3072,0.019759999381171334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,4096,0.04030133287111918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,3584,0.014144889182514615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,3584,0.029647111892700195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,3584,0.037273777855767146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,3072,0.013095999757448832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,3072,0.02613777750068241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,2048,0.014396443963050842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,3072,0.03371466530693902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,2560,0.012031111452314587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,2560,0.02271644439962175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,1536,0.011234666738245221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,2560,0.029548443026012842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,2048,0.011039999624093374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,2048,0.01924088928434584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,1024,0.008829333715968663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,2048,0.025975111458036635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,1536,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,1536,0.01605866683853997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,768,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,1536,0.021393777595625982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,1024,0.009197333620654212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,1024,0.012774222426944308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,6144,0.045540445380740695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,512,0.005978666659858491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,1024,0.017846221725145977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,768,0.008974221845467886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,768,0.011169777976142036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,256,0.004550222307443619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,4096,0.03284710976812575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,512,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,768,0.01622488929165734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,512,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,2560,0.016358221570650738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,256,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,512,0.014442667365074158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,256,0.008833777573373582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,64,0.0038497779104444715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,32,0.004218666503826777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,128,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,256,0.013063111239009433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,128,0.008319999608728621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,128,0.010461333725187512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,64,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,32,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,65536,0.27727556228637695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,51200,0.20587555567423502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,65536,0.1528257793850369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,16384,0.07402044534683228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,51200,0.12111377716064453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,65536,0.39320532480875653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,51200,0.30778490172492134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,12288,0.05819733275307549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,16384,0.04713155494795906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,16384,0.09988533125983344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,65536,0.4807226922776964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,10240,0.04745066828197903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,51200,0.37664087613423664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,16384,0.1329937775929769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,12288,0.037611554066340126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,12288,0.07432978020773993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,2048,128,0.004523555437723796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,12288,0.09900533490710789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,10240,0.028836445675955877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,10240,0.06308977471457587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,7168,0.03232622146606445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,8192,0.02274133265018463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,8192,0.051444444391462535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,6144,0.02696977721320258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,8192,0.07034844160079956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,7168,0.02071377800570594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,7168,0.045863999260796436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,5120,0.024130667249361675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,6144,0.018647111124462552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,6144,0.040241777896881104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,6144,0.05539910991986593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,5120,0.034663110971450806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,4096,0.019688000281651814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,5120,0.01679111189312405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,5120,0.04758666621314155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,3584,0.017265778448846605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,4096,0.01475288967291514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,4096,0.028970665401882593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,8192,0.04048177931043837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,4096,0.04010844561788771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,3072,0.01553066737122006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,3584,0.013809777796268463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,3584,0.026510222090615168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,2560,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,3584,0.03654488921165466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,3072,0.012855999999576144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,3072,0.02314044369591607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,2048,0.01162755572133594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,3072,0.0322693321439955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,2560,0.011904000408119626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,2560,0.02085955606566535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,7168,0.062069336573282875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,1536,0.008866666919655269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,2560,0.028347555134031508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,2048,0.010757333702511258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,2048,0.017281777328915067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,1024,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,2048,0.025054223007626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,1536,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,1536,0.014481777946154276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,768,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,1536,0.02128977742460039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,1024,0.009214222431182861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,512,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,1024,0.01147555559873581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,768,0.008626666333940294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,256,0.0041777777175108595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,768,0.010062221851613786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,512,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,768,0.015895111693276297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,128,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,512,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,512,0.014427555931939019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,256,0.00797155582242542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,256,0.007807110746701558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,64,0.0035013332962989807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,10240,0.08443555566999648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,256,0.012755555411179861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,128,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,128,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,128,0.01035911093155543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,64,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,65536,0.18687200546264648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,32,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,51200,0.13291200002034506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,65536,0.10768889056311713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,16384,0.04916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,65536,0.2993768850962321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,51200,0.08625688817765977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,65536,0.3375875684950087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,51200,0.2357849015129937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,12288,0.038076443804634943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,1024,0.01756266587310367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,16384,0.029133333100212946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,51200,0.26643111970689565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,10240,0.03050666716363695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,16384,0.09500622087054783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,12288,0.023570666710535686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,12288,0.05906133519278633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,8192,0.02511022157139248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,10240,0.021212443709373474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,12288,0.07319644424650404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,10240,0.050776888926823936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,7168,0.021272889441914026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,10240,0.061794665124681264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,8192,0.01650311052799225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,8192,0.041132443481021456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,6144,0.019137778215938143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,8192,0.052072889275021024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,7168,0.015145777000321282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,7168,0.03712977634535895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,5120,0.01642311116059621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,7168,0.04673333300484551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,6144,0.01390044391155243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,6144,0.0325617790222168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,4096,0.013384888569513956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,6144,0.04134933484925164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,5120,0.027875555886162653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,5120,0.03602755400869582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,4096,0.011409777734014722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,16384,0.07861689064237806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,4096,0.029606223106384277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,3584,0.0211208893193139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,3584,0.010761777559916178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,3072,0.01068800025516086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1536,32,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,3072,0.010394666757848527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,2560,0.009561777942710454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,3072,0.019056888090239633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,3072,0.025243555506070454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,2560,0.009700444837411245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,2560,0.016789333687888253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,2048,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,2560,0.022357333037588332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,5120,0.012478222449620565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,1536,0.0068915556702348925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,2048,0.009192888935407003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,2048,0.014454222387737699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,2048,0.019711999429596793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,1024,0.005535111245181825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,1536,0.012113778127564324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,1536,0.0086986662613021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,1536,0.017179555363125272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,1024,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,768,0.0047662220895290375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,1024,0.009937778115272522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,1024,0.014495111174053617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,512,0.003944888710975647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,768,0.007608888877762689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,768,0.008908444808589088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,768,0.01311555587583118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,256,0.0034977777136696708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,512,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,512,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,512,0.0124177775449223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,256,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,256,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,256,0.011368888947698804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,4096,0.023212444451120164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,64,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,128,0.0069857777820693115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,32,0.0028320000403457214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,128,0.008728000024954477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,65536,0.15962933169470891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,64,0.006568888823191325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,32,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,3584,0.027590221828884546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,51200,0.12670844131045872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,65536,0.10632444752587213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,65536,0.2590880129072401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,16384,0.043226665920681424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,51200,0.2055964469909668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,65536,0.33444621827867294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,128,0.0028826666788922418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,12288,0.02791822287771437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,51200,0.26595645480685765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,16384,0.02871911062134637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,16384,0.06791644626193576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,10240,0.024675556355052527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,1024,3584,0.012345777617560493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,12288,0.02330933345688714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,12288,0.05127999848789639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,8192,0.01889866590499878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,10240,0.020971554848882887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,10240,0.04386311107211643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,10240,0.06116533279418945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,7168,0.017450667089886136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,8192,0.01635199950801002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,8192,0.03613511059019301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,7168,0.032574223147498235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,51200,0.08492355876498753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,6144,0.015584000282817416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,7168,0.015132443772421943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,6144,0.013814222481515674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,5120,0.01348444405529234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,6144,0.02842666705449422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,6144,0.04058666692839728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,16384,0.09424799680709839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,4096,0.011387555963463254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,128,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,5120,0.012717333104875354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,5120,0.024658666716681585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,3584,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,5120,0.0351893338892195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,4096,0.011393778026103973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,4096,0.02909244431389703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,3072,0.009291555318567488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,3584,0.010709332923094431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,3584,0.018924444913864136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,8192,0.05116800136036343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,2560,0.008623999853928884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,3072,0.01035822182893753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,3072,0.016924444172117446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,7168,0.04599377844068739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,3072,0.024579554796218872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,2048,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,2560,0.009698666632175446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,2560,0.014897776974572076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,2560,0.02199555602338579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,1536,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,2048,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,2048,0.01941244469748603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,12288,0.07122755712933011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,1536,0.01094133324093289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,1024,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,4096,0.020555555820465088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,1536,0.008656889200210571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,1536,0.016656888855828177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,768,0.004439110971159405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,1024,0.008347555167145198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,1024,0.009138666921191746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,512,0.003547555456558863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,3584,0.027440889014138117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,1024,0.014436443646748861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,768,0.008286221987671321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,768,0.013109332985348172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,256,0.0035093331502543557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,512,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,512,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,128,0.003157333367400699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,512,0.012054222325483957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,256,0.007099555598364935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,256,0.006325333482689328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,256,0.011030221978823343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,64,0.0030337776988744736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,128,0.0069822221994400024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,768,32,0.0032115555885765287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,2048,0.013101333545313941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,128,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,65536,0.10789866579903497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,128,0.008647110727098253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,64,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,51200,0.08784000078837077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,65536,0.10398755470911662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,65536,0.2175840006934272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,16384,0.03323377834426032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,51200,0.08390311400095622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,51200,0.17198310958014595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,65536,0.32317066192626953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,768,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,12288,0.02421599957678053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,16384,0.027408000495698717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,51200,0.25552354918585884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,16384,0.057010663880242236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,10240,0.020245333512624104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,12288,0.022203556365436975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,16384,0.0902551081445482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,8192,0.016899555921554565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,12288,0.04387288954522875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,10240,0.020269332660569083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,12288,0.06935111019346449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,7168,0.01510933372709486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,10240,0.03779200050565932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,8192,0.015781333049138386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,10240,0.059076441658867725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,6144,0.013261333107948303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,8192,0.03129511078198751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,7168,0.01460800071557363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,8192,0.0494391123453776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,7168,0.02809777855873108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,6144,0.013433777623706393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,768,32,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,4096,0.00997688869635264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,6144,0.024732444021436904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,6144,0.03853155506981744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,5120,0.012095111111799875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,5120,0.02125599980354309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,3584,0.009027555584907532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,5120,0.03329599897066752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,4096,0.011043555206722684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,3072,0.0083155557513237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,4096,0.017978666557206046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,4096,0.028258668051825628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,3584,0.010480889015727572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,3584,0.016591111818949383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,2560,0.007353777686754863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,3584,0.02606755495071411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,3072,0.010057777994208867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,3072,0.014829332629839579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,2048,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,3072,0.023531556129455566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,2560,0.009556444154845344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,2560,0.013580444786283704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,1536,0.005464000006516774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,2560,0.021478222476111516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,2048,0.008989333278603023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,1024,0.004252444538805219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,2048,0.018755555152893066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,1536,0.008377778033415476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,5120,0.011504000259770287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,1536,0.00977244476477305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,768,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,1536,0.01622222198380364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,7168,0.04391466577847799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,1024,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,1024,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,512,0.0035413333939181436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,768,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,1024,0.013793778088357715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,768,0.007381333245171442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,256,0.002874666617976295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,768,0.013082666529549493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,512,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,512,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,128,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,512,0.011839110818174152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,256,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,256,0.005969777703285217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,64,0.002191999927163124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,128,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,512,32,0.0025111111915773815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,256,0.010748444332016839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,128,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,65536,0.06174933248096042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,128,0.008270222279760573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,64,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,51200,0.051705777645111084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,32,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,512,2048,0.011699555648697747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,65536,0.10370044575797187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,65536,0.173927108446757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,16384,0.017774222625626456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,51200,0.08367111285527547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,12288,0.017138666576809354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,16384,0.026660444007979497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,65536,0.3181920051574707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,16384,0.046484443876478404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,10240,0.01566133399804433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,51200,0.2514764467875163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,16384,0.08814666668574016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,12288,0.02146222194035848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,8192,0.012930666406949362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,12288,0.036159111393822566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,10240,0.01996799972322252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,10240,0.030954665607876245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,7168,0.011868444581826528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,10240,0.05789688560697767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,8192,0.015526221858130561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,8192,0.048476444350348584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,6144,0.010327999790509542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,7168,0.014496889379289416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,7168,0.02296977738539378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,5120,0.008991999758614434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,7168,0.04304800099796719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,6144,0.013295110729005603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,6144,0.020407110452651978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,4096,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,6144,0.03756977783309089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,5120,0.012111110819710625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,5120,0.017993777990341187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,51200,0.13630666997697619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,3584,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,5120,0.03247822324434916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,4096,0.010974222587214576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,3072,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,4096,0.015415110521846347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,4096,0.02762311034732395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,3584,0.01425866617096795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,2560,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,12288,0.06776711013582018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,3584,0.0255395554833942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,3072,0.010063110954231685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,3072,0.012857777376969656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,2048,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,3072,0.02313422163327535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,8192,0.025606221622890894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,1536,0.004223111189073986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,2560,0.011595555477672152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,2048,0.009017777939637503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,2560,0.020720889170964558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,2048,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,1024,0.0034924443397257063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,2048,0.01850844422976176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,1536,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,1536,0.008839999636014303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,1536,0.016120000018013846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,1024,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,1024,0.007595555649863348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,512,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,1024,0.013451555536852943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,768,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,768,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,256,0.0025111111915773815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,512,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,512,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,3584,0.0103493332862854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,512,0.010999999940395355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,256,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,256,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,256,0.01035644445154402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,128,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,64,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,2560,0.009418666362762451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,128,0.005212444398138259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,32,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,128,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,64,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,256,32,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,768,0.003181333343187968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,51200,0.042360888587103955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,65536,0.17284977436065674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,768,0.012712000144852532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,65536,0.31666845745510525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,51200,0.13549778196546766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,16384,0.011954666839705573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,256,128,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,51200,0.08323022392061022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,51200,0.24960711267259386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,12288,0.012768000364303589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,16384,0.04605866803063286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,16384,0.08690933386484782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,65536,0.051670223474502563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,12288,0.02134311033619775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,12288,0.035232888327704534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,12288,0.06715644730461968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,8192,0.012301333248615265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,10240,0.019653333557976615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,10240,0.030226667722066242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,7168,0.011071110765139261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,65536,0.10389510790506999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,8192,0.0155102229780621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,10240,0.057152887185414634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,6144,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,8192,0.02509511179394192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,8192,0.0466231107711792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,7168,0.014303111367755465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,5120,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,7168,0.022669333550665114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,6144,0.013221333424250284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,7168,0.0417795545525021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,6144,0.020120888948440552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,4096,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,6144,0.03700977894994948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,16384,0.026035555534892615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,5120,0.017647110753589206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,3584,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,5120,0.03237333231502109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,4096,0.01073155552148819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,4096,0.015210666590266757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,3072,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,10240,0.011426666544543372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,4096,0.02737422287464142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,3584,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,2560,0.0052844443255000645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,3584,0.014117333624098035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,3584,0.02533600065443251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,3072,0.012773333324326409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,2048,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,2560,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,3072,0.022967999180157978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,1536,0.003909333298603694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,2560,0.011624000138706632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,2560,0.02054666644997067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,2048,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,1024,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,2048,0.010059555371602377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,2048,0.018183110488785636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,1536,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,768,0.0028906667398081887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,1536,0.015834665960735746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,1024,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,1024,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,512,0.0026213334252436957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,1024,0.013398222625255585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,768,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,768,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,256,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,768,0.012447111308574677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,512,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,512,0.010801778071456485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,128,0.0021031111892726687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,256,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,256,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,3072,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,256,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,64,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,128,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,128,32,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,128,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,65536,0.04708088768853081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,128,0.00776711106300354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,51200,0.03875289029545254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,16384,0.010693333215183683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,64,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,128,32,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,1536,0.00832799987660514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,10240,0.010650667051474253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,65536,0.17309599452548555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,8192,0.009143999881214565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,16384,0.045887112617492676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,12288,0.03478577733039856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,7168,0.008395555946562026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,5120,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,10240,0.029923554923799302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,6144,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,5120,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,512,0.007090667055712805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,8192,0.024902222885025874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,4096,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,7168,0.022658665974934895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,3584,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,3072,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,6144,0.019938665959570143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,2560,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,5120,0.017549332645204332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,2048,0.004524444540341695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,4096,0.0151582227812873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,1536,0.0038888889054457345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,3584,0.014122666584120857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,1024,0.0031964443624019623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,3072,0.012728888955381183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,768,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,2560,0.011360000405046674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,512,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,2048,0.010098666780524785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,256,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,12288,0.012096888489193387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,128,0.002157333410448498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,1024,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,64,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,768,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,512,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,51200,0.1356711122724745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,256,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,65536,0.04744977752367655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,128,0.005295111073387994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,64,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,51200,0.03564000129699707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,16384,0.009756444229020013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,32,0.005109333329730564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,12288,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,10240,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,65536,0.1723946730295817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,16384,0.04555199874771965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,51200,0.13524977366129556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,12288,0.03472977876663208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,7168,0.010226666927337646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,6144,0.009541333549552495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,10240,0.029904001288943823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,5120,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,4096,0.007904888855086433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,8192,0.024846222665574815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,7168,0.022533333963818018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,3584,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,6144,0.019987555013762582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,5120,0.017613333132531907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,64,1536,0.0087013335691558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,2560,0.005062222066852781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,2048,0.004388444539573458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,3584,0.014108444253603617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,64,32,0.0018666666001081467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,1536,0.0037208890749348533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,1024,0.003170666595300039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,2560,0.011424889167149862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,768,0.0028239999794297745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,2048,0.010085333552625444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,512,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,1536,0.009024000002278222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,256,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,1024,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,128,0.0018915554715527429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,768,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,64,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,512,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,8192,0.011032000184059143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,32,0.0018728888697094387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,256,0.005294222384691238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,128,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,64,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,32,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2048,32,3072,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,4096,0.015153777268197803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2048,32,3072,0.012802666260136498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,16384,0.79202757941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,12288,1.7649431228637695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,16384,2.309075461493598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,12288,0.6159795655144585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,12288,1.9648994869656031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,16384,2.5682150522867837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,10240,0.5026684337192112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,10240,1.4380088382297094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,8192,1.2025751537746854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,51200,2.8454596201578775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,10240,1.532019509209527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,8192,0.4002391232384576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,8192,1.1756132973564994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,7168,1.002621332804362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,7168,0.35894399219089085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,6144,0.31397512223985463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,6144,0.921812481350369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,7168,1.0438604354858398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,6144,0.9307840135362414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,5120,0.7180995411343045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,4096,0.6831822395324707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,5120,0.7586702240837945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,4096,0.2367182307773166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,4096,0.5808284547593858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,3584,0.5559564696417915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,3584,0.19604710737864176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,3584,0.510408878326416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,5120,0.26310756471421987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,3072,0.460768010881212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,51200,7.312728034125434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,3072,0.17331200175815156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,2560,0.40350845124986434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,2560,0.16149244043562147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,3072,0.48609246148003477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,2560,0.36720532841152614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,2048,0.3415022161271837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,2048,0.14851733048756918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,2048,0.3064071072472466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,1024,0.19239377975463867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,768,0.15856266021728516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,1024,0.08153688907623291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,1024,0.23551021681891549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,1536,0.23860621452331543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,512,0.09999999735090469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,256,0.06460177898406982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,768,0.14401511351267496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,512,0.06640622350904676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,128,0.05567110909356011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,512,0.12627200285593668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,768,0.08835733599132961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,256,0.06045599778493246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,256,0.11127821604410808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,64,0.06996088557773165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,128,0.05830933650334676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,32,0.053732444842656456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,65536,1536,0.23532178666856554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,128,0.10520622465345596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,64,0.10578933027055527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,65536,1536,0.10253866513570149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,32,0.1082035568025377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,65536,51200,10.272078620062933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,16384,1.769426663716634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,16384,0.6017973158094618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,51200,2.239323510064019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,16384,2.0176177554660373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,65536,2.905259450276693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,12288,1.4254621929592555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,12288,0.4770088725619846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,12288,1.4772621790568035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,10240,0.3916782273186578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,10240,1.3225955963134766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,51200,5.660253312852647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,10240,1.2298302120632596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,8192,0.33308532502916127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,8192,1.0223076078626845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,8192,0.9988364113701714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,7168,0.8007600042555066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,7168,0.2832426759931776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,65536,7.2282316419813375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,7168,0.823728879292806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,6144,0.85025421778361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,6144,0.24520089891221789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,5120,0.6429528660244411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,6144,0.6863333384195963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,5120,0.2102559937371148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,4096,0.4813520113627116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,4096,0.17322933673858643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,3584,0.3843546708424886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,4096,0.459046204884847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,51200,8.10896979437934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,3584,0.42485955026414657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,3584,0.15394311481051975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,3072,0.3831520080566406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,3072,0.1370346678627862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,2560,0.29723644256591797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,2560,0.11968089474572076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,3072,0.3505946795145671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,2048,0.2650079992082384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,2560,0.2910364468892415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,2048,0.11541867256164551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,2048,0.2371537685394287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,1024,0.14024799399905735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,1536,0.08512266476949056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,1536,0.19272177749209932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,768,0.10978133148617214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,1024,0.0643048882484436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,1024,0.15495110882653132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,512,0.08685333198971218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,768,0.05832711193296644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,768,0.11727733082241482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,5120,0.5806684494018555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,512,0.054584887292650014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,512,0.09915822082095677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,128,0.04511733187569512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,256,0.04990399877230326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,64,0.03903911179966397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,256,0.08826400174034967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,32,0.040467556979921125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,128,0.08368088801701863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,1536,0.17910756005181205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,64,0.08381688594818115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,32,0.08460978004667495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,51200,65536,10.737758212619356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,51200,256,0.056490666336483426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,51200,128,0.04673511121008131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,65536,0.8716275427076551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,16384,0.5950506528218588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,51200,0.6599066522386339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,51200,2.0401670667860246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,16384,0.20070932971106636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,12288,0.48965957429673934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,16384,0.7012862099541558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,12288,0.15463911162482366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,65536,2.5580755869547525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,10240,0.1336524486541748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,10240,0.43807376755608457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,8192,0.32597067621019155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,12288,0.4681119918823242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,10240,0.3763937685224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,8192,0.10910311010148789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,7168,0.29499732123480904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,7168,0.0988817744784885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,51200,2.4894951714409723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,6144,0.24242401123046875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,65536,3.306072023179796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,5120,0.19600266880459258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,7168,0.3080586592356364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,6144,0.09240888886981541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,6144,0.22898401154412162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,4096,0.16349955399831137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,4096,0.0603662199444241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,3584,0.05496711201137967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,3072,0.12695466147528753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,5120,0.1957119968202379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,3584,0.13317067093319362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,2560,0.0987502204047309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,3072,0.047817776600519814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,2048,0.07807555463578966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,3072,0.11961955494350857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,2560,0.04376088910632663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,2560,0.10001866685019599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,1536,0.06039200226465861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,2048,0.03755733370780945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,8192,0.31477154625786674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,1024,0.04045777850680881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,2048,0.0835919976234436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,1536,0.03005511230892605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,768,0.031898667414983116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,1536,0.06627644432915582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,5120,0.0741164419386122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,512,0.025790222816997107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,768,0.022456889351209004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,1024,0.050092445479498975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,256,0.017765333255132038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,768,0.04065955678621928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,3584,0.13699200418260363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,128,0.013964444398880005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,512,0.0203840004073249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,64,0.012067555553383298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,256,0.01868266695075565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,4096,0.15551288922627768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,16384,32,0.012367111113336353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,256,0.029996444781621296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,128,0.028903110159768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,64,0.02846933404604594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,32,0.028484443823496502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,1024,0.024898666474554274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,65536,0.6847893397013346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,16384,512,0.03329866793420579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,16384,128,0.018578666779730055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,51200,1.403331544664171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,51200,0.5443413522508409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,65536,1.8315821753607855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,16384,0.5889582104153103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,16384,0.4693653318617079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,12288,0.3801697889963786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,12288,0.12163110574086507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,10240,0.351931545469496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,51200,1.896665785047743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,12288,0.39008000161912704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,10240,0.28535734282599556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,8192,0.27127467261420357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,7168,0.21786933475070527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,8192,0.29064443376329213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,7168,0.07519555754131742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,6144,0.18455199400583902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,7168,0.2011653317345513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,16384,0.16228532791137695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,6144,0.06953244739108615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,5120,0.15440977944268122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,5120,0.057358225186665855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,4096,0.1393724415037367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,5120,0.14619199434916177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,3584,0.11928621927897136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,4096,0.04743644595146179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,4096,0.12727377149793836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,3072,0.10610666539933945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,3584,0.043522665898005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,3584,0.10330755843056573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,10240,0.12242044342888726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,3072,0.03784622086419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,8192,0.09012177917692397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,2560,0.07617600096596612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,2048,0.06844800048404269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,3072,0.09217155641979641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,2560,0.03333688775698344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,1536,0.04610133171081543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,2560,0.07771022452248467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,2048,0.029071112473805744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,1024,0.032604446013768516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,2048,0.06477422184414335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,768,0.025729777084456548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,1536,0.05157599846522013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,1024,0.020198222663667467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,512,0.01942311061753167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,1024,0.039410667286978826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,768,0.01850755512714386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,256,0.014130666851997375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,768,0.03231822119818793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,512,0.01664622293578254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,128,0.011008000208271874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,512,0.026269333230124578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,256,0.015493333339691162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,64,0.00963199966483646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,256,0.023346667488416035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,12288,32,0.01001422190003925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,128,0.015490666031837463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,128,0.022642667094866436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,64,0.022662222385406494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,65536,2.6048995123969183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,32,0.022658665974934895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,12288,1536,0.024338665935728285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,12288,6144,0.18345155980851915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,65536,0.5591813193427192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,51200,0.45000531938340926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,16384,0.4299982123904758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,16384,0.13565599918365479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,51200,1.2151048448350694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,65536,1.5676790873209636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,12288,0.33695822291904026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,12288,0.10391644636789958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,10240,0.27521154615614146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,12288,0.3113635645972358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,10240,0.10074488984213935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,51200,1.5003467135959203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,8192,0.20952088303036162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,8192,0.07304266426298353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,65536,2.0025181240505643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,7168,0.19494844807518852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,7168,0.06401333543989393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,8192,0.21898133224911162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,6144,0.05835555659400093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,5120,0.12063554922739665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,6144,0.1716195609834459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,4096,0.094014220767551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,5120,0.050307555331124194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,5120,0.12837066915300158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,3584,0.09206222163306342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,4096,0.03988800115055508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,4096,0.10269777642356025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,3584,0.035952889257007174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,16384,0.4554497930738661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,3584,0.09129155344433254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,2560,0.06341689162784152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,3072,0.07767199807696872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,2560,0.029475556479560003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,10240,0.24957688649495444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,2560,0.06640977991951837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,1536,0.039408889081743025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,6144,0.17074044545491537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,2048,0.025049777494536504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,7168,0.17705244488186309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,2048,0.055757330523596875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,1536,0.021218667427698772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,768,0.023057777020666335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,1536,0.0444124440352122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,1024,0.017983999517228868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,512,0.016159999701711867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,1024,0.03428533342149522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,768,0.01621955633163452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,768,0.0284204449918535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,512,0.014817777607176038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,3072,0.0766471094555325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,128,0.009653333160612319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,256,0.013829333086808523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,3072,0.0328382220533159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,64,0.008593777815500895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,256,0.020561777883105807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,32,0.008760888543393876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,10240,128,0.013798221945762634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,128,0.019705777366956074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,2048,0.05203288793563843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,64,0.019838222199016146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,32,0.019648000597953796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,1024,0.027673777606752183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,65536,0.48276445600721574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,10240,256,0.014560000763999091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,10240,512,0.022841778066423204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,51200,0.37834845648871523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,65536,1.1088790893554688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,16384,0.3275573253631592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,51200,1.0690995322333443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,16384,0.11559466520945232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,16384,0.31920888688829213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,10240,0.19994400607215035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,12288,0.11176533169216579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,51200,1.1937120225694444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,65536,1.508639971415202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,12288,0.2699786557091607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,10240,0.0739226672384474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,10240,0.19795733027988008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,7168,0.14075911045074463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,8192,0.1818151076634725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,6144,0.11733155780368382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,7168,0.05318311187956068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,5120,0.10153066449695164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,6144,0.04867555697758993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,6144,0.12200000551011826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,5120,0.040435555908415056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,4096,0.07938310835096571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,5120,0.10239644183052911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,12288,0.22188622421688506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,3584,0.0647217763794793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,4096,0.03335288829273648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,4096,0.08436266581217448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,3584,0.02975289026896159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,3584,0.07333066728379992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,8192,0.18949333826700845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,3072,0.027095999982621934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,2560,0.04525688952869839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,3072,0.06425511174731784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,2048,0.03820800118976169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,2560,0.024550222688251074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,1536,0.030772444274690416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,2560,0.05578311284383138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,2048,0.02163733376397027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,7168,0.14106667041778564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,1536,0.018716444571812946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,768,0.01856355534659492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,1536,0.03726933399836222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,1024,0.015606222881211175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,512,0.013549333645237817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,1024,0.028806222809685603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,768,0.014313777287801107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,256,0.013962666193644205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,768,0.023945776952637568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,512,0.013261333107948303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,128,0.00793688909875022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,3072,0.059233778052859835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,256,0.012487110992272695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,64,0.006896889044178857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,256,0.017263111140992906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,32,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,128,0.012242666549152799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,128,0.01684977776474423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,8192,8192,0.059225777784983315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,64,0.01688088807794783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,32,0.01684266659948561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,8192,1024,0.022664000590642292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,2048,0.04650311006440056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,65536,0.41985421710544163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,16384,0.29571466975741917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,51200,0.3324151039123535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,8192,512,0.019370666808552213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,51200,0.8773991266886393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,16384,0.10378755463494195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,12288,0.2093600034713745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,10240,0.19189155101776123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,12288,0.09708711173799302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,16384,0.3492622110578749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,12288,0.20760177241431343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,65536,1.3104719585842557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,51200,1.0700150595770943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,8192,0.15003377861446804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,10240,0.07920266522301568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,7168,0.12283999390072292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,10240,0.19285778204600015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,8192,0.05400177836418152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,6144,0.11388533645206028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,7168,0.054057776927948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,5120,0.10210577646891277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,8192,0.1442613336775038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,6144,0.04372533493571811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,7168,0.13461866643693712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,4096,0.08111644453472562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,5120,0.04429866539107429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,6144,0.1136053270763821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,4096,0.031572444571389094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,5120,0.1001431081030104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,3072,0.05971911218431261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,4096,0.07407822211583455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,65536,0.9659289254082574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,3584,0.0343431101904975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,3584,0.06504444281260173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,3072,0.025807110799683466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,3072,0.05808089176813761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,2560,0.023403555154800415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,1536,0.028210666444566514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,2048,0.02034222251839108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,2560,0.04982577761014303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,2048,0.041322665082083807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,1024,0.02015555567211575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,1536,0.017509332961506315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,768,0.0162435554795795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,1536,0.03395377927356296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,1024,0.014542222023010254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,512,0.012437333663304647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,1024,0.02567377852069007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,768,0.01349422252840466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,256,0.00903911143541336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,768,0.021888888544506494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,512,0.012522666818565793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,128,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,512,0.017508443858888414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,3584,0.06624444325764973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,256,0.011720000041855706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,256,0.015853333804342482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,64,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,7168,128,0.011649777491887411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,2560,0.044841776291529335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,128,0.015457777513398064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,2048,0.0358222226301829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,64,0.015475556254386902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,7168,32,0.01520088811715444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,65536,0.3713671101464166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,51200,0.2937528822157118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,16384,0.24103111690945098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,51200,0.8771004676818848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,65536,0.9958373175726997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,51200,0.9150951173570422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,16384,0.2561955451965332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,7168,32,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,12288,0.17293955220116508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,65536,1.1506569120619032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,10240,0.17197688420613608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,12288,0.0759377810690138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,12288,0.18906488683488634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,8192,0.12915555636088052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,7168,0.10130755768881904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,10240,0.15368710623847112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,8192,0.05137688914934794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,7168,0.04447022080421448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,6144,0.0855404469701979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,8192,0.12490310933854844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,5120,0.08212000131607056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,7168,0.10933066738976373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,4096,0.0682613319820828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,5120,0.03352888756328159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,6144,0.09542755285898845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,3584,0.06103377872043186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,4096,0.02864355511135525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,4096,0.06510844495561388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,3072,0.05205511053403219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,3584,0.025601777765485976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,3584,0.057806220319536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,2560,0.043339557117886014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,3072,0.023397333092159692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,16384,0.10917688740624322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,2048,0.03273155623012119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,3072,0.05123733480771383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,2560,0.021264889174037512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,1536,0.02595111065440708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,2560,0.04411910971005758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,2048,0.018613333503405254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,10240,0.07035199801127116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,1024,0.017749332719379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,768,0.013533333109484779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,1536,0.015863999724388123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,1024,0.01388711151149538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,512,0.010750222537252637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,768,0.012902222573757172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,1024,0.022908444205919903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,256,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,6144,0.039679110050201416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,512,0.011769777370823754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,128,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,512,0.01590222285853492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,256,0.014436443646748861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,5120,0.08032888836330838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,64,0.006221333311663733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,6144,32,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,128,0.013800889253616333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,64,0.013819555441538492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,32,0.013743110828929476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,2048,0.03675555520587497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,1536,0.030066665675905015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,65536,0.3284328778584798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,6144,768,0.01960177719593048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,65536,0.8281377686394585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,256,0.011085333095656501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,16384,0.20873510837554932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,51200,0.6105759938557943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,6144,128,0.011160888605647616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,51200,0.25291289223564994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,16384,0.07852977514266968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,12288,0.16264443927341038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,65536,0.9543804592556424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,10240,0.13388178083631727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,12288,0.05983289082845052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,8192,0.10365333159764607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,10240,0.052079998784595065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,51200,0.7557617823282877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,12288,0.1699742211235894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,7168,0.09349777963426377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,10240,0.13073333104451498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,8192,0.04292444388071696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,6144,0.07847288582060072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,8192,0.10596444209416707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,7168,0.04042933384577433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,5120,0.06488355663087633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,6144,0.03464622298876444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,7168,0.10117510954538982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,4096,0.052391111850738525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,3584,0.045053333044052124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,6144,0.08119822210735746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,4096,0.026124444272783067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,5120,0.06907911433113946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,3072,0.041145778364605375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,3584,0.023189332750108507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,4096,0.05791733662287394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,2560,0.035467555125554405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,3584,0.05269955595334371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,3072,0.021592888567182753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,2048,0.026231999198595684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,3072,0.04487644301520454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,1536,0.02016444504261017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,2560,0.03903999924659729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,2048,0.01683911184469859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,1024,0.014855111638704935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,16384,0.26623556349012584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,2048,0.032720890310075544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,1536,0.014448000325096978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,768,0.012131555212868584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,1024,0.012508444488048553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,1536,0.026734222968419392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,512,0.009138666921191746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,1024,0.020392888122134738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,256,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,512,0.010760000182522668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,768,0.01719111038578881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,128,0.006194666855865055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,512,0.014312888185183207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,5120,0.029847111966874864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,256,0.01035377797153261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,256,0.012890666723251343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,64,0.0053591111467944244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,128,0.01035288886891471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,5120,32,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,128,0.012381333443853589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,64,0.012422222230169507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,5120,32,0.012321777641773224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,2560,0.018928888771269057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,65536,0.27943287955390084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,65536,0.5945475366380479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,51200,0.47270578808254665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,51200,0.22784267531500924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,16384,0.14949333667755127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,5120,768,0.011726222104496427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,16384,0.06698489189147949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,12288,0.13142489062415227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,10240,0.09978399674097697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,65536,0.7844319873385959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,12288,0.061343113581339516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,16384,0.19789332813686797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,10240,0.06093600061204699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,12288,0.1286835538016425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,7168,0.06689244508743286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,8192,0.044365333186255566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,10240,0.1108364462852478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,6144,0.05851733022265964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,8192,0.09620800283220078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,5120,0.05039822061856588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,6144,0.030468444029490154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,7168,0.0842648877037896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,4096,0.04084355632464091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,5120,0.026288888520664636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,6144,0.07159733110004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,3584,0.03357244531313578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,5120,0.059318224589029946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,4096,0.02197955548763275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,3072,0.02951288885540432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,4096,0.049201776583989464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,3584,0.020145777199003432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,2560,0.026139555705918208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,3584,0.043476445807351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,2048,0.021528000632921856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,3072,0.03806844353675842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,2560,0.01665422154797448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,1536,0.01720622181892395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,2560,0.03356000118785434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,2048,0.014763555592960782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,1024,0.012792888614866467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,51200,0.5989635785420736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,2048,0.028185778194003638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,1536,0.013422222601042854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,768,0.011062222222487131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,1536,0.02241155505180359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,1024,0.011596444580290051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,512,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,768,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,1024,0.017394666870435078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,7168,0.0324577788511912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,768,0.015138667490747241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,256,0.006903999795516332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,128,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,512,0.010009778042634329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,512,0.01242844429281023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,256,0.009448889229032729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,64,0.004562666846646203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,256,0.01146399974822998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,128,0.009367999931176504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,32,0.004854222138722737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,128,0.011015999648306104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,64,0.010843555960390302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,4096,32,0.010743111371994019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,65536,0.4702231089274089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,65536,0.25819378428988987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,4096,8192,0.08185866806242201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,51200,0.4341199927859836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,16384,0.1406995587878757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,12288,0.10090488857693142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,65536,0.687064011891683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,51200,0.20276178254021537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,16384,0.0797377758555942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,16384,0.1549395587709215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,10240,0.09205955266952515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,12288,0.0495760010348426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,8192,0.071578668223487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,12288,0.12338756190405951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,10240,0.11247821648915608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,8192,0.03484355409940084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,7168,0.06386755572424994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,6144,0.047819554805755615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,8192,0.08189777533213298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,4096,3072,0.018607111440764535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,7168,0.03166844447453817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,5120,0.04528533418973287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,6144,0.028427554501427546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,7168,0.0732204450501336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,4096,0.0368115570810106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,5120,0.024643555283546448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,6144,0.06264977984958224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,3584,0.03211288981967502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,5120,0.05362400081422594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,4096,0.021458667185571458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,3072,0.02831288841035631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,4096,0.043248888519075185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,3584,0.019519110520680744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,2560,0.022791999909612868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,3584,0.03881422347492642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,2048,0.018479110466109384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,2560,0.016344888342751395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,1536,0.01534577707449595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,2048,0.014589332871966891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,2048,0.024872000018755596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,10240,0.0433208876185947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,1536,0.012798222402731577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,1536,0.0206675562593672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,768,0.009218666288587783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,1024,0.011400000088744693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,1024,0.016159999701711867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,512,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,768,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,768,0.014078223043017916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,256,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,512,0.009856888817416297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,512,0.01144444445768992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,128,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,256,0.009301332963837517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,256,0.010718222293588849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,3072,0.017858665850427415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,64,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3584,128,0.009298666483826106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,3072,0.03464711043569777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,32,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,2560,0.030095110336939495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,64,0.010054222411579555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3584,1024,0.011219555305110084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,32,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,65536,0.21544799539777967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,65536,0.5191742049323188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,51200,0.3844008975558811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,51200,0.17452711529201934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,16384,0.11833333969116211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,12288,0.08888889021343654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,65536,0.5990853309631348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,16384,0.05749599801169502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,51200,0.4690853224860297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,10240,0.07894577582677205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,16384,0.15841510560777453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,12288,0.04680622286266751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,12288,0.10531644688712226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,10240,0.038169778055614896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,7168,0.055177778005599976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,8192,0.0324942237801022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,10240,0.09013688564300537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,51200,0.538525316450331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,6144,0.047840889957216054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,8192,0.0738240016831292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,7168,0.02749155627356635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,5120,0.03635199864705404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,6144,0.026905778381559584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,7168,0.06454666455586751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,4096,0.032045334577560425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,6144,0.0564791096581353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,3584,0.02790044413672553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,5120,0.04803555541568332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,4096,0.019305777218606737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,3072,0.023990222149425085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,3584,0.017083555459976196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,3584,0.03569066524505615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,2560,0.02023555503951179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,3072,0.015571556157535978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,3072,0.031727999448776245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,2048,0.01646933290693495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,8192,0.06366311179267035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,2560,0.027789334456125896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,1536,0.01366133325629764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,2048,0.013184000220563678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,2048,0.023188443647490606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3584,128,0.01033688916100396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,1024,0.009323555562231276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,1536,0.011530666715568967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,768,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,1024,0.01036177741156684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,1536,0.018983110785484314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,512,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,1024,0.014997333288192749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,5120,0.022435555855433147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,256,0.005106666849719154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,768,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,128,0.0046977777447965415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,512,0.009124444590674506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,768,0.013022222452693515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,256,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,512,0.010699555277824402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,256,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,128,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,128,0.009317333499590555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,64,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,32,0.009381333159075843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,3072,2560,0.014420444766680399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,65536,0.36457154485914445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,65536,0.20089244842529297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,51200,0.32361777623494464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,16384,0.10544533199734157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,51200,0.15776533550686306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,65536,0.510457780626085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,12288,0.07918044593599108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,51200,0.3994319968753391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,16384,0.051629334688186646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,10240,0.06477333439721002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,12288,0.04028622309366862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,16384,0.12295377254486084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,64,0.004181333300140169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,8192,0.052130666044023305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,10240,0.03624533282385932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,7168,0.04689777890841166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,10240,0.0768266651365492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,6144,0.041068444649378456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,8192,0.06495822138256498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,7168,0.05654399924808078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,6144,0.024504888388845656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,5120,0.03293155630429586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,6144,0.04865244362089369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,4096,0.02747022277779049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,5120,0.020769778225156996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,5120,0.041887111133999295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,4096,0.01793155570824941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,3584,0.02388444377316369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,4096,0.03436977664629618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,3072,0.020302222834693063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,3584,0.016580444243219163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,3584,0.031112889448801678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,3072,4096,0.03961955507596334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,12288,0.09320088889863755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,3072,0.015053333507643806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,3072,32,0.004202666795916027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,3072,0.027639110883076985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,8192,0.03018310997221205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,2560,0.013766222529941134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,1536,0.01239999963177575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,7168,0.02604177759753333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,2048,0.01274400038851632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,1024,0.009142222503821055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,2048,0.02015199926164415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,1536,0.01124533348613315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,768,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,1536,0.01704355577627818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,1024,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,512,0.006340444501903322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,1024,0.01349244432316886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,768,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,256,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,768,0.011667555405033959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,128,0.0041893331540955436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,512,0.009947555760542551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,256,0.009099555512269339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,128,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,128,0.00869777798652649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,2560,0.017646221650971305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,2048,0.015196444259749519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,32,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,64,0.008678221868144141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,2560,0.02381155557102627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2560,32,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,65536,0.3349946604834662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,65536,0.1568791071573893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,51200,0.2568559911515978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,512,0.008957333034939235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,16384,0.08120710982216729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2560,256,0.008364444805516137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,51200,0.12490399678548177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,12288,0.06050399939219157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,51200,0.3475493325127496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,16384,0.04130311144722833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,10240,0.05039822061856588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2560,64,0.0038897775941424896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,16384,0.10406133201387192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,12288,0.03289066751797994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,8192,0.04132711225085788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,7168,0.03576711151334975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,10240,0.03033155534002516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,12288,0.08027733034557767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,8192,0.02346222268210517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,10240,0.06911200284957886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,8192,0.05526666508780586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,7168,0.022494221727053326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,5120,0.026830222871568467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,6144,0.019339554839664035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,7168,0.049398223559061684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,4096,0.020384889509942796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,6144,0.043362667163213096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,5120,0.017080000705189176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,3584,0.018326222896575928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,5120,0.03935822182231479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,4096,0.015214223000738354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,3072,0.016737777325842116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,4096,0.031191110610961914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,3584,0.013952889376216464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,2560,0.014127999544143677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,3584,0.02831733226776123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,65536,0.4437582227918837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,2048,0.011762667033407422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,3072,0.02500444485081567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,1536,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,2560,0.021807110971874658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,2048,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,1024,0.007511999871995714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,2048,0.018375999397701688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,1536,0.009895111123720804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,768,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,1536,0.015511999527613321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,1024,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,512,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,1024,0.012431999875439538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,6144,0.030425777037938435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,256,0.0041751112374994485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,512,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,128,0.003536000019974179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,512,0.009208888643317753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,256,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,64,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,256,0.008299555215570662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,2048,32,0.0035342222286595237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,128,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,64,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,3072,0.013103110922707452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,65536,0.25710754924350315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,2560,0.011696889168686338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,51200,0.19841421975029838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,65536,0.14971555603875056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,16384,0.06226666768391927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,65536,0.3366542127397325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,768,0.008640888664457533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,51200,0.1340284479988946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,51200,0.2646755642361111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,16384,0.038478222158220075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,12288,0.0481333335240682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,10240,0.042071110672420926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,16384,0.08714310990439521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,12288,0.03149244520399306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,8192,0.030763556559880573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,10240,0.02830577890078227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,2048,128,0.007702222300900354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,10240,0.05533866749869453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,8192,0.022600889205932617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,6144,0.024079999989933435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,8192,0.04542311032613119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,7168,0.020305777589480083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,32,0.007660444411966536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,5120,0.020218667056825426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,7168,0.04074755642149184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,4096,0.016136889656384785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,6144,0.035951998498704694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,5120,0.016176000237464905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,3584,0.01514488955338796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,5120,0.030646221505271062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,4096,0.014783110883500842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,4096,0.025916443930731878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,3072,0.01297688898113039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,3584,0.01366488883892695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,3584,0.023131555981106226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,2560,0.011742221812407175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,2048,768,0.010671999719407825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,3072,0.012720888687504662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,3072,0.020488000578350492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,2560,0.011487111449241638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,2560,0.018021333548757765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,1536,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,2048,0.010403555300500657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,2048,0.015492444237073263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,1024,0.006607999818192587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,1536,0.009638222555319468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,1536,0.013103110922707452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,6144,0.018381334013409086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,768,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,1024,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,512,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,1024,0.010703999963071613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,768,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,256,0.003855111284388436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,768,0.009248889154858058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,512,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,128,0.0031662223239739737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,256,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,2048,0.009555555880069733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,64,0.0031866667171319327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,128,0.007332444190979004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,12288,0.07271199756198458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,7168,0.028561777538723413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,64,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,65536,0.21346488263871935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,65536,0.10131111409929063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,51200,0.16089599662356907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,512,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,65536,0.28526579009162056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,51200,0.08875111076566909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,16384,0.051889777183532715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,256,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,12288,0.039889779355790876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,51200,0.22395200199551055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,16384,0.029245333539115057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,10240,0.03241422110133701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,16384,0.07310133510165744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,12288,0.02242044442229801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1536,128,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,8192,0.02703022294574314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,10240,0.02091911103990343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,7168,0.02368444369898902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,12288,0.05567466550403171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,10240,0.0491866668065389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,8192,0.0166640000210868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,7168,0.014869333969222175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,8192,0.039035555389192365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,5120,0.01830311119556427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,6144,0.014144000079896716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,7168,0.035210665729310774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1536,32,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,4096,0.01534844438234965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,6144,0.030991110536787245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,5120,0.012432888978057437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,4096,0.011367999845080905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,5120,0.02629599968592326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,3072,0.011527111132939657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,3584,0.01070844464831882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,4096,0.022114665971861944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,2560,0.01035822182893753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,3584,0.01995466649532318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,3072,0.010320000350475311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,3072,0.018227555685573153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,2560,0.016192000773217943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,1536,0.007349333001507654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,2048,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,2048,0.013789333403110504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,1024,0.005764444255166584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,1536,0.008599999878141616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1536,32,0.0031946665710873077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,1536,0.011738667057620155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,768,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,6144,0.02050844497150845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,512,0.004528000122971005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,1024,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,1024,0.00959911114639706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,768,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,256,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,512,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,768,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,128,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,3584,0.013746666411558786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,64,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,256,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,128,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,32,0.002946666752298673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,128,0.0064604443808396654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,1024,2048,0.008519111408127679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,64,0.006344888773229387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,65536,0.1364204486211141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,2560,0.009707555174827576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,32,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,51200,0.11638489034440781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,65536,0.0994044476085239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,16384,0.03860355416933695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,51200,0.0803697771496243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,65536,0.24078578419155547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,12288,0.0273297776778539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,16384,0.027172444595230952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,51200,0.18978844748602974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,10240,0.023152000374264185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,16384,0.06254755788379245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,12288,0.02236533330546485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,8192,0.018795554836591084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,12288,0.04844355583190918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,10240,0.020584000481499564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,7168,0.01682933337158627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,10240,0.041257778803507485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,8192,0.01661422186427646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,1024,512,0.0074684446056683855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,6144,0.015298666225539314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,8192,0.033909334076775446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,7168,0.030639999442630347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,6144,0.01368977791733212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,4096,0.011065777805116443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,6144,0.026761778526835974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,5120,0.01239822225438224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,5120,0.023346667488416035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,3584,0.009993777506881291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,4096,0.011192888849311404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,3072,0.008908444808589088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,4096,0.019664888580640156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,3584,0.017788444956143696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,2560,0.007800888684060838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,3072,0.016152888536453247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,2560,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,2560,0.0143857780430052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,1024,256,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,7168,0.01519555515713162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,1536,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,2048,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,2048,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,1024,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,1536,0.00870222184393141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,768,0.00387999994887246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,1024,0.008016000191370646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,1536,0.0107351111041175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,1024,0.008800889054934183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,512,0.003526222374704149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,768,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,3584,0.01069155583779017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,256,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,512,0.007448888487286038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,768,0.007911110917727152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,128,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,3072,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,256,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,64,0.0028284444577164124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,768,128,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,32,0.0028115556471877625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,256,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,128,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,65536,0.10964088969760472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,2048,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,64,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,51200,0.08911910984251235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,16384,0.03336355421278212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,65536,0.09881777895821466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,768,5120,0.01293688929743237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,51200,0.07963022258546618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,12288,0.028332445356580947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,16384,0.02607111136118571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,51200,0.15940354930029974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,10240,0.023160000642140705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,16384,0.0529075562953949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,12288,0.02141155633661482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,8192,0.018780445059140522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,12288,0.04119199845525954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,10240,0.02010933392577701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,7168,0.01702577703528934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,10240,0.035183111826578774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,8192,0.01551999979548984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,6144,0.015207111835479736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,8192,0.029037333197063867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,7168,0.014471999473041959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,5120,0.013343111508422427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,512,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,7168,0.026496888862715826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,6144,0.013069333301650153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,4096,0.010671999719407825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,3584,0.009757333331637913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,5120,0.011792889071835412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,4096,0.01073155552148819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,5120,0.02022133270899455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,3072,0.008614222208658854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,3584,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,4096,0.017216889394654166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,2560,0.007698666718271043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,3584,0.015541333291265698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,3072,0.009840889109505547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,2048,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,3072,0.014175999495718213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,65536,0.2047662205166287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,1536,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,2560,0.012651555240154266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,2048,0.008949333594905006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,2048,0.011025778121418424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,1024,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,1536,0.008329778081840938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,768,0.004239999999602635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,1536,0.009376000199053023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,1024,0.007690666450394525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,512,0.0035342222286595237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,1024,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,768,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,256,0.0028604444944196274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,768,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,512,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,512,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,128,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,6144,0.023324444890022278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,64,0.002472888885272874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,256,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,128,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,512,32,0.0024622221373849446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,768,32,0.006154666758245892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,128,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,65536,0.06496089034610324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,64,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,51200,0.052876445982191295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,2560,0.009288888838556077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,512,32,0.00562755556570159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,16384,0.018242667118708294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,65536,0.09850311279296875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,12288,0.015255999233987598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,51200,0.07938755883110894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,65536,0.1580008930630154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,16384,0.024958221448792353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,10240,0.017144888639450073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,51200,0.12391555309295654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,16384,0.04201777776082357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,8192,0.012770666844315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,12288,0.020988444487253826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,7168,0.011350221931934357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,10240,0.01960177719593048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,12288,0.032736000087526106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,6144,0.010008888940016428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,10240,0.028204444381925795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,7168,0.014203555054134794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,8192,0.023319111929999456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,5120,0.00905777762333552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,512,256,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,6144,0.013067555924256643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,4096,0.007521777517265744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,6144,0.018958222534921434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,5120,0.012139555480745105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,5120,0.01665866706106398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,4096,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,4096,0.014482667048772177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,3072,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,3584,0.010320000350475311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,2560,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,3584,0.013385777672131857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,3072,0.009777777724795872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,2048,0.004640000147951974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,2560,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,3072,0.012080888781282636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,1536,0.004233777936961916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,2560,0.01093066649304496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,2048,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,1024,0.0035128887328836652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,2048,0.009717333647939894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,1536,0.00831822223133511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,768,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,1024,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,1536,0.00833777752187517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,8192,0.015354666444990369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,512,0.0027955555253558685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,256,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,768,0.00738844441043006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,1024,0.0075706665714581805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,128,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,512,0.0070222218831380205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,768,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,7168,0.021330666210916307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,256,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,512,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,32,0.0021840000732077495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,256,128,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,256,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,65536,0.046896000703175865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,128,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,51200,0.038000000847710505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,64,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,256,32,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,16384,0.015104000767072042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,51200,0.07882222202089098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,65536,0.15649867057800293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,12288,0.012628444366984896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,16384,0.02460622290770213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,51200,0.12321689393785264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,10240,0.010623999767833285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,16384,0.041747556792365186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,12288,0.021145777569876775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,8192,0.008387555678685507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,12288,0.03249955508444045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,10240,0.019309333629078336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,7168,0.009288888838556077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,10240,0.027637332677841187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,8192,0.015390222271283468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,6144,0.00867022242810991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,8192,0.023220444718996685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,7168,0.014096000128322177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,5120,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,64,0.0021155555215146807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,7168,0.021179555190934077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,6144,0.01314755529165268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,4096,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,6144,0.018931556079122756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,256,3584,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,5120,0.011789333489206104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,4096,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,5120,0.016544888416926067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,3072,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,65536,0.09825066725413005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,4096,0.014319111903508505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,3584,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,2560,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,3584,0.01313066648112403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,3072,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,2560,0.009337777892748514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,3072,0.012053333222866058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,1536,0.00387999994887246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,2560,0.011035555766688453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,2048,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,1024,0.0032088889016045462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,2048,0.009647111097971598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,1536,0.008360000120268928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,768,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,1536,0.00868888861603207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,1024,0.007691555553012424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,512,0.002529777793420686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,1024,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,768,0.007333333293596904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,256,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,768,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,512,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,128,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,256,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,512,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,64,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,32,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,256,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1793,128,128,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,65536,0.043134222428003945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,128,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,51200,0.03360355562633938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,16384,0.01236177815331353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,32,0.005197333378924264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,12288,0.011025778121418424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,2048,0.004522666749027041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,65536,0.1562524504131741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,10240,0.009707555174827576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,8192,0.009248889154858058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,16384,0.04135555691189236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,12288,0.03195555673705207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,6144,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,10240,0.02770400047302246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,8192,0.023021333747439917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,5120,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,4096,0.00675733346078131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,7168,0.020727111233605277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,3584,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,6144,0.01869422197341919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,3072,0.00554044461912579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,5120,0.01648177703221639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,2560,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,4096,0.014296889305114746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,3584,0.013096888860066732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,2048,0.004578666554556953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,1536,0.003896000070704354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,128,3584,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,3072,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,1024,0.003198222153716617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,128,64,0.005087999833954705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,768,0.0028799999919202593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,2560,0.010747555229398938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,512,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,256,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,2048,0.009641778137948778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,1536,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,128,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,1024,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,32,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,51200,0.12303377522362603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,768,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,512,0.006081777728266186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,64,7168,0.008607111043400234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,65536,0.0392506652408176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,256,0.005535999933878581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,51200,0.03164266546567281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,16384,0.009369778136412302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,128,0.005117333183685939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,64,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,10240,0.007663999994595845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,64,32,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,8192,0.006904888898134232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,65536,0.15687821971045599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,16384,0.04121422105365329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,7168,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,12288,0.031889776388804116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,6144,0.009639999932712978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,10240,0.02719199988577101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,5120,0.009392889009581672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,4096,0.00813155538505978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,8192,0.023016888234350417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,3584,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,7168,0.02087466584311591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,6144,0.018571555614471436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,5120,0.016523554921150208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,4096,0.014440889159838358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,2560,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,2048,0.005020444591840108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,3072,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,1536,0.0038559999730851916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,2560,0.010880889164076911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,2048,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,1024,0.0031964443624019623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,768,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,12288,0.008264888491895463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,512,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,1024,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,768,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,256,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,128,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,512,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,64,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,256,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,128,0.005035555611054103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,32,0.001820444439848264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1793,32,3072,0.006506666541099548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,64,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,3584,0.013186666700575085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,32,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,1536,0.008603555460770925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1793,32,51200,0.12264177534315322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,16384,0.6793786684672037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,16384,2.3175760904947915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,16384,2.4852044847276473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,12288,1.73815737830268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,16384,2.120011647542318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,12288,1.7862826453314886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,12288,0.5230213271247016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,51200,2.593137741088867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,12288,1.6397590637207031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,10240,1.367023997836643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,10240,0.45801687240600586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,10240,1.4820213317871094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,51200,6.96831554836697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,8192,0.3754302130805121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,8192,1.1215812895033095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,8192,1.1336338255140517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,10240,1.3033395343356662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,7168,0.3201457659403483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,8192,1.0283803939819336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,7168,0.9427892896864148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,7168,1.0235084957546658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,6144,0.8206853336758084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,6144,0.28196001052856445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,5120,0.6627937952677408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,7168,0.9254977968004016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,6144,0.8461466895209419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,5120,0.7383431328667535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,5120,0.2449235651228163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,51200,9.612837897406683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,4096,0.5710666444566515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,6144,0.7711866696675619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,4096,0.20486933655209014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,4096,0.5520586437649196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,5120,0.6204240057203505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,3584,0.5176124572753906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,3584,0.1826524469587538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,3072,0.44393687778049046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,3584,0.5136933326721191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,4096,0.5069022178649902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,3072,0.16212356090545654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,3072,0.44228532579210067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,3584,0.4556408988104926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,2560,0.3828133212195502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,2560,0.14098666773902044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,2048,0.27620087729560006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,2560,0.35968534151713055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,2560,0.3350248866611057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,1536,0.2079448832405938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,2048,0.31420887841118705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,2048,0.11962132983737522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,1024,0.14906044801076254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,1536,0.10600622495015462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,2048,0.3826879925198025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,1536,0.23327112197875977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,1024,0.17831110954284668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,1536,0.2261111206478543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,1024,0.08587644497553508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,768,0.10643021927939521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,1024,0.16883733537462023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,512,0.07895200120078193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,768,0.06940088669459026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,768,0.1407359970940484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,768,0.13773955239189997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,256,0.05976088841756185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,512,0.06286933024724324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,512,0.1221911112467448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,512,0.11060000128216213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,128,0.05018488897217644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,256,0.056333336565229625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,256,0.10812177922990586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,65536,128,0.053710222244262695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,64,0.046367999580171376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,128,0.10196177826987372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,65536,32,0.04645511176851061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,128,0.06708533234066434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,64,0.10296799739201863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,65536,32,0.10294755299886067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,3072,0.39356266127692324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,256,0.08712355295817058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,65536,51200,7.137396918402778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,65536,2.6276302337646484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,51200,1.9699598948160808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,51200,5.328386518690321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,16384,2.1581634945339627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,65536,7.0643971761067705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,16384,0.5482648743523492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,16384,2.083645290798611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,12288,1.1401341756184895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,51200,7.690572102864583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,12288,1.4973857667711046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,12288,0.4117591116163466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,16384,1.6195546256171331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,10240,1.262420442369249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,12288,1.208539538913303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,10240,0.35324443711174863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,65536,10.395523071289062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,10240,1.2348488701714408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,8192,0.28984088367886013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,51200,5.352873908148871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,8192,0.8976418177286783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,8192,0.9189466900295682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,10240,0.976586659749349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,7168,0.8290115462409126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,7168,0.2663431167602539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,8192,0.7793413268195258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,7168,0.8278311093648275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,6144,0.7235448625352648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,6144,0.2275644408331977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,6144,0.7348462210761176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,5120,0.5313617918226454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,7168,0.6634382141960992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,5120,0.19233600298563638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,65536,7.107584211561416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,5120,0.5636204613579644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,4096,0.41399643156263566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,4096,0.16176711188422307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,6144,0.6086737844679091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,3584,0.4065413210127089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,4096,0.4629386795891656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,3584,0.14422044489118788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,5120,0.5318053563435873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,3584,0.38678399721781415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,4096,0.3928435643513997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,3072,0.33699999915228945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,3072,0.12855377462175158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,2560,0.28100087907579213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,3584,0.36189688576592344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,2048,0.24640088611178926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,3072,0.42195823457505965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,2560,0.28265510665045845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,2560,0.25935912132263184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,2048,0.26536443498399526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,2048,0.09976266490088569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,3072,0.31018487612406415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,2048,0.21698221895429823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,1024,0.11286132865481907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,1536,0.18376978238423666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,1536,0.07686399751239352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,768,0.08775822321573894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,1024,0.06357510884602864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,1024,0.1405093272527059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,1536,0.1721386644575331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,512,0.06483999888102214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,768,0.11248800489637588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,768,0.10930844148000081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,512,0.04983911249372694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,256,0.05000444584422641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,512,0.09069955348968506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,2560,0.14014489120907253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,128,0.03804355528619554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,256,0.045239110787709556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,1536,0.17328977584838867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,256,0.08597777949439155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,256,0.07031110922495525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,64,0.03256266646915012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,128,0.04320889049106174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,128,0.0811048878563775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,51200,32,0.034127109580569796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,128,0.05429955654674106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,64,0.08123022317886353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,32,0.0817813343471951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,51200,768,0.05527377790874905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,51200,512,0.09640177753236558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,65536,0.8158079783121744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,51200,1024,0.13282310962677002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,51200,1.9741573333740234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,65536,2.5422000885009766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,51200,0.6292479832967123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,16384,0.5517768859863281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,16384,0.2088693380355835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,65536,3.1534514957004123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,51200,2.465661366780599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,16384,0.616595533159044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,12288,0.4632710880703396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,12288,0.14384710788726807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,16384,0.5361830923292372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,10240,0.3413173357645671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,12288,0.4624951150682237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,10240,0.12215733528137207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,8192,0.2820231119791667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,10240,0.3886088795132107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,12288,0.3958408832550049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,7168,0.20402310954199898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,8192,0.10551289055082534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,10240,0.3305271201663547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,65536,2.4183057149251304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,7168,0.09367644124560887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,6144,0.22242667939927843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,7168,0.25419288211398655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,6144,0.24682399961683485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,7168,0.23681065771314833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,5120,0.19072443909115264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,51200,1.9488951365152996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,5120,0.06715999709235297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,4096,0.15362400478786892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,6144,0.20622666676839194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,5120,0.18875555197397867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,5120,0.1749955548180474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,3584,0.11912622716691758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,4096,0.05707199705971611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,4096,0.15034222602844238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,8192,0.2969760100046794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,4096,0.15135378307766384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,3072,0.11705244912041558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,3584,0.1298560036553277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,8192,0.27343644036187065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,2560,0.07398933172225952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,3072,0.04509955644607544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,3072,0.11386399798923069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,2048,0.08267644378874037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,6144,0.08425244357850815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,2560,0.0406968891620636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,3072,0.11460267172919379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,2560,0.09812000062730576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,1536,0.059713780879974365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,2048,0.034972445832358465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,2560,0.09750488731596206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,2048,0.08101511001586914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,1024,0.035750223530663386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,2048,0.08351022005081177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,1536,0.02831911047299703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,1536,0.06503466765085857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,768,0.034837332036760114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,1536,0.06845066944758098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,1024,0.02347022294998169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,1024,0.04929244518280029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,512,0.02481599979930454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,768,0.0210826661851671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,1024,0.05338577760590447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,256,0.019880000087949965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,3584,0.0512586666478051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,768,0.03950400153795878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,768,0.04566577739185757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,512,0.01902933253182305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,128,0.01479822231663598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,512,0.032511999209721885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,3584,0.12819733884599474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,512,0.03878666626082526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,64,0.012492444780137805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,256,0.02924799919128418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,128,0.017255999975734286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,256,0.030838221311569214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,128,0.02808977829085456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,16384,128,0.023631999890009563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,64,0.027997334798177082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,16384,32,0.02797599964671665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,65536,0.6072319878472222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,16384,256,0.01754133403301239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,16384,32,0.012704888979593912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,51200,1.4328799777560766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,16384,0.509692456987169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,51200,0.4803848796420627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,51200,1.6138328976101344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,65536,2.119500478108724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,16384,0.15227466159396702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,16384,0.4657617674933539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,65536,2.466417736477322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,12288,0.4050613244374593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,16384,0.4252435631222195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,12288,0.11365155378977458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,12288,0.34711289405822754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,10240,0.30023643705579967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,12288,0.3078320026397705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,10240,0.09508444203270806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,8192,0.2435795466105143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,10240,0.2904951042599148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,10240,0.26211733288235134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,7168,0.20252088705698648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,8192,0.08700622452629937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,8192,0.22294844521416557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,8192,0.21173422866397432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,7168,0.07509333557552762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,6144,0.16772177484300402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,51200,1.414377742343479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,7168,0.19887644714779326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,5120,0.13766400019327799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,6144,0.06052888764275444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,7168,0.22655023468865287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,6144,0.169250660472446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,4096,0.11703288555145264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,6144,0.1622355514102512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,5120,0.056667553053961865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,5120,0.15635466575622559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,3584,0.09516888856887817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,4096,0.04502222273084852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,4096,0.12081955538855658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,65536,1.8978631761338975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,4096,0.11480000283983018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,3584,0.03987555702527364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,3584,0.10319111082288955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,2560,0.07238844368192884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,3072,0.03590133455064561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,3072,0.08962133195665146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,2048,0.055607113573286265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,3072,0.09144088957044814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,2560,0.07679555813471477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,2560,0.03156800071398417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,1536,0.044822222656673856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,2560,0.07889066802130805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,2048,0.027781334188249376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,2048,0.06280622217390272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,1024,0.027951111396153767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,2048,0.06748355759514703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,1536,0.02254666719171736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,1536,0.05070222086376614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,768,0.022551111049122278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,1536,0.0553653339544932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,1024,0.01942488882276747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,1024,0.03854577740033468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,512,0.016736888223224215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,1024,0.04389511214362251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,768,0.03176533182462057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,768,0.03769866625467936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,3584,0.10372533400853474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,256,0.01833066675398085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,512,0.015994666351212394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,512,0.025734222597546045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,128,0.015352000792821249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,512,0.03188622328970168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,256,0.01464355488618215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,256,0.022679999470710754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,64,0.01408266690042284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,256,0.025570667452282373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,128,0.015009777413474189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,128,0.022061333060264587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,128,0.019139554765489366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,64,0.022055110997623865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,12288,32,0.02201866606871287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,12288,5120,0.13806666268242732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,12288,768,0.017857778403494094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,65536,0.5105359819200304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,32,0.014372444815105863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,65536,1.4409137301974828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,51200,1.00891908009847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,16384,0.37059911092122394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,12288,3072,0.07776533232794867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,51200,0.4059448772006565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,65536,2.073298772176107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,16384,0.12258222368028428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,51200,1.5289991166856554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,16384,0.39666223526000977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,12288,0.29018667009141713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,12288,0.09811555676990086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,10240,0.24994754791259766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,16384,0.35206932491726345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,12288,0.3306471241845025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,8192,0.1870373355017768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,10240,0.08186399936676025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,10240,0.2654862138960096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,65536,1.5144746568467882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,7168,0.16835999488830566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,8192,0.06667288806703356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,10240,0.21926400396558973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,8192,0.19013422065311006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,51200,1.1147759755452473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,7168,0.059847997294531934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,6144,0.1449822187423706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,7168,0.18512711260053846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,6144,0.052969776921802096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,6144,0.16909866862826875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,5120,0.11547288629743789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,6144,0.13723733690049914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,4096,0.0960551102956136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,5120,0.04577689038382637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,5120,0.11918311648898655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,5120,0.1244000064002143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,4096,0.03752622339460585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,3584,0.0828311112191942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,4096,0.09747555520799424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,4096,0.09814844528834026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,3072,0.06478933493296306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,3584,0.03362577822473314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,7168,0.1958773268593682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,8192,0.18203643957773843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,3584,0.08858666817347209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,2560,0.061611553033192955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,3072,0.030758221944173176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,3072,0.0757039984067281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,3072,0.07777955796983507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,2048,0.04539111256599426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,2560,0.0653688907623291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,2560,0.02756711178355747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,1536,0.03435111045837402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,2560,0.06735644737879436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,2048,0.023968888653649226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,1024,0.023736889163653057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,2048,0.05387644304169548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,12288,0.2560311158498128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,2048,0.05765066544214884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,1536,0.043415112627877124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,768,0.018939556346999276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,1536,0.04895288745562235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,1024,0.016933333542611863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,1024,0.033611555894215904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,512,0.01459111107720269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,1024,0.03803999887572394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,768,0.015280889140235053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,768,0.027658666173617046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,256,0.010966222319338055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,768,0.032227555910746254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,512,0.014137778017255994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,512,0.022543110781245764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,512,0.02778844369782342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,256,0.013182222015327878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,256,0.019837333096398246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,64,0.008841777841250101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,256,0.021665778425004747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,32,0.008982222113344405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,128,0.013448889056841532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,128,0.019294222195943195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,10240,128,0.017460443907313876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,64,0.01926399932967292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,32,0.019259555472267997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,10240,1536,0.02067466742462582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,10240,3584,0.08597066667344834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,10240,128,0.010237333675225576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,65536,1.2079138225979276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,51200,0.9429119957817925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,51200,0.3423013422224257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,16384,0.3314969009823269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,65536,1.5958924823337133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,51200,1.1955288781060112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,16384,0.3293946584065755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,16384,0.10515466663572524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,12288,0.25093422995673287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,65536,0.42200801107618546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,12288,0.08357155323028564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,16384,0.3080977863735623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,12288,0.2663031154208713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,10240,0.2097733285692003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,65536,1.2735315958658855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,12288,0.22738933563232422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,8192,0.16772266228993735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,10240,0.06901777452892728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,51200,1.0088959799872506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,7168,0.14241600036621094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,8192,0.05533688929345873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,8192,0.17589688301086426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,10240,0.1935537788603041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,8192,0.16859822803073457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,7168,0.05219733383920458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,7168,0.13636177115970188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,5120,0.10894933011796738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,6144,0.04439466529422336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,6144,0.13506578074561224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,7168,0.14880088965098062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,6144,0.12184977531433105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,5120,0.03808800048298306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,5120,0.10708710882398818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,4096,0.08102755414115058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,4096,0.03246577911906772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,3584,0.07413422399097018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,4096,0.08619911140865749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,3584,0.02955199943648444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,3584,0.07122933202319674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,3072,0.06475022104051378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,10240,0.18937155935499403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,3584,0.07812800010045369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,2560,0.051560001240836255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,6144,0.1313004493713379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,3072,0.026856889327367146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,3072,0.06332533227072822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,3072,0.06872177786297269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,2560,0.024040000306235418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,2560,0.05460355679194132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,4096,0.08571466472413804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,2560,0.06042844719356961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,5120,0.09956088993284438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,2048,0.020788444413079154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,1536,0.03197244471973843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,2048,0.045075555642445884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,1024,0.022621333599090576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,2048,0.051691555314593844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,1536,0.017418666018380057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,768,0.021000888612535264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,1536,0.03683999843067593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,1024,0.014870221416155497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,1024,0.02829155657026503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,512,0.016508445143699646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,768,0.013712888790501488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,768,0.023575110567940608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,768,0.029037333197063867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,256,0.01219733307758967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,512,0.01908177799648709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,512,0.01262399968173769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,128,0.010433778166770935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,512,0.024246222443050806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,2048,0.04063733418782552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,256,0.019944889677895438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,64,0.009295110901196798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,128,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,128,0.016542222764756944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,8192,32,0.010000000397364298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,128,0.016354666815863717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,64,0.016506666938463848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,32,0.016341333587964375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,1536,0.04296000136269463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,8192,1024,0.03394488824738397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,65536,0.38218755192226833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,8192,256,0.011786667009194693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,8192,256,0.016921778519948322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,65536,1.1675795449150932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,51200,0.7382799784342448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,16384,0.2771751085917155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,51200,0.30509866608513725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,65536,1.382847997877333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,16384,0.09194222423765396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,51200,1.0342701805962455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,16384,0.3678453233506944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,12288,0.20678044690026176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,16384,0.27514221933152944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,12288,0.07738844553629558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,10240,0.1798755592770047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,12288,0.22277156511942545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,65536,1.067207972208659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,8192,0.143068446053399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,10240,0.06362222300635444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,10240,0.18273154894510904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,7168,0.13125422265794542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,8192,0.056073778205447726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,8192,0.1380924383799235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,8192,0.15245689286126032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,6144,0.10671199692620172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,7168,0.05023555623160469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,7168,0.12422845098707412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,5120,0.08983733256657918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,7168,0.12522044446733263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,6144,0.040207998620139226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,6144,0.10553066598044501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,5120,0.035135110219319664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,6144,0.10884088940090603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,4096,0.06707021925184461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,5120,0.08924177620146011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,3584,0.0641991098721822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,5120,0.0940577785174052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,4096,0.02977422210905287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,4096,0.07231555382410686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,4096,0.07810933060116239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,12288,0.20179466406504312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,3584,0.026610665851169165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,3584,0.06404978036880493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,10240,0.17636177274915907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,2560,0.04307555490069919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,3072,0.023968888653649226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,51200,0.8747688929239908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,2048,0.03565422362751431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,3072,0.056856890519460045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,2560,0.02187999917401208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,2560,0.04871999886300829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,1536,0.02676711148685879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,2560,0.054409778780407377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,2048,0.019136000010702345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,2048,0.04085244403945075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,2048,0.046945777204301625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,1024,0.020416888925764296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,1536,0.016772444049517315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,1536,0.03319199879964193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,1536,0.03887822230656942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,768,0.016234666109085083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,1024,0.013996443814701505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,1024,0.025651555922296312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,1024,0.03125422199567159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,3072,0.05464977688259549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,512,0.012434666355450949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,768,0.01312444441848331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,3584,0.07043644454744127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,256,0.009740444521109263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,512,0.012051555845472546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,512,0.017058667209413316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,512,0.02247200078434414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,3072,0.06304533614052667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,256,0.011389333340856763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,256,0.015286222100257874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,256,0.01867022282547421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,64,0.007627555893527136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,7168,128,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,128,0.01482222146458096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,32,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,128,0.015449777245521545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,64,0.014879999889267815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,32,0.014827556080288358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,7168,768,0.021310221817758348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,7168,768,0.026681777503755357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,65536,0.3364231056637234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,7168,128,0.008626666333940294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,65536,1.074435551961263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,51200,0.7833093537224664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,16384,0.2955253389146593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,51200,0.8850124677022299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,65536,0.9363066355387369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,16384,0.27108976576063365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,12288,0.16952978240119088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,16384,0.08829244640138413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,16384,0.2542648845248752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,10240,0.16866489251454672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,51200,0.7309573491414388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,12288,0.20594222015804717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,12288,0.17921777566274008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,10240,0.05963022179073758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,8192,0.13230578104654947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,65536,1.1422835456000435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,51200,0.2809075514475505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,10240,0.15295199553171793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,8192,0.045215109984079994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,7168,0.12002221743265788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,8192,0.12184000015258789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,6144,0.1002417802810669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,8192,0.12649688455793592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,7168,0.03981155488226149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,7168,0.1055377788013882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,7168,0.10921155744128758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,6144,0.03592977921168009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,6144,0.09257866938908894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,4096,0.06585688723458184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,5120,0.030995554394192163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,5120,0.07747822337680392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,12288,0.06390488810009427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,5120,0.08208088742362128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,3584,0.05700622002283732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,4096,0.025780444343884785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,10240,0.15166044235229492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,3072,0.04831022355291578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,3584,0.023413333627912734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,3584,0.05673155519697401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,3584,0.06224089198642307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,2560,0.040976001156700984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,3072,0.05063733458518982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,3072,0.05555377735031975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,5120,0.08420888582865398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,2048,0.030814223819308813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,2560,0.043783111704720386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,2560,0.0195777780479855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,2560,0.04910044537650215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,1536,0.023471110396915015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,6144,0.09497955772611831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,2048,0.017143110434214275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,2048,0.03667555583847894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,1024,0.016587555408477783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,2048,0.041821334097120494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,1536,0.014774221513006421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,4096,0.06390400065316094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,1536,0.02996444371011522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,768,0.013925333817799887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,1536,0.035164445638656616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,1024,0.01293155550956726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,4096,0.06988977723651461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,512,0.01089244418674045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,1024,0.022482666704389784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,1024,0.02794933319091797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,768,0.01204533295498954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,768,0.01934400035275353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,256,0.008048888709810045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,512,0.011155555645624796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,512,0.015512888630231222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,3072,0.021577777134047613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,128,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,256,0.01036800030204985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,512,0.020516445239384968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,256,0.014105778601434497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,64,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,256,0.017856889300876193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,6144,32,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,128,0.013516444298956128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,128,0.014241778188281588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,64,0.013524444566832649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,6144,32,0.013468444347381592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,65536,0.6104328897264268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,65536,0.2823866738213433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,51200,0.5424070888095432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,6144,768,0.02421066661675771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,65536,0.9332835939195422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,16384,0.18733510706159803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,6144,128,0.01037688884470198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,51200,0.2139164474275377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,12288,0.13316533300611708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,16384,0.06839111116197374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,65536,0.7723804579840766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,51200,0.7432497872246636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,10240,0.11693599489000107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,16384,0.1977004475063748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,12288,0.053145776192347206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,51200,0.6290115780300564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,12288,0.15038932694329157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,8192,0.0982968873447842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,10240,0.04594044552909004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,12288,0.15736977259318033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,10240,0.13214311334821913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,7168,0.08543644348780315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,8192,0.03683200147416856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,10240,0.1342746681637234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,8192,0.10457066694895427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,6144,0.07165244552824232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,7168,0.033300442828072443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,7168,0.09245510896046956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,5120,0.05980088975694445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,6144,0.029848890172110662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,7168,0.09479377667109172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,6144,0.08014844523535834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,4096,0.04647022154596117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,6144,0.08148266871770223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,5120,0.026425777210129633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,5120,0.06817511055204603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,3584,0.04279377725389269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,5120,0.07100888755586413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,4096,0.023024888502226934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,16384,0.21883732742733428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,4096,0.06060444646411472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,3584,0.020791111720932853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,3584,0.050322665108574756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,2560,0.02993866801261902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,3584,0.05381066600481669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,3072,0.019036443697081674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,3072,0.04404888881577385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,2048,0.023387556274731953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,2560,0.01719466679626041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,2560,0.03858133488231235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,2560,0.04234222239918179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,8192,0.10709510909186469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,1536,0.019187556372748483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,2048,0.015224888920783997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,2048,0.03208533260557387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,1024,0.01387466655837165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,2048,0.03632355398601956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,1536,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,1536,0.026535999443795946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,1536,0.03091644578509861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,1024,0.01146577795346578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,1024,0.020104888412687514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,1024,0.024159111910396155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,3072,0.03734933336575826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,512,0.009007111191749573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,768,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,768,0.01707466608948178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,256,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,512,0.010037333601050908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,512,0.014000889327791003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,512,0.018800000349680584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,128,0.006335999816656113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,256,0.012602667013804117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,256,0.009630222287442949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,256,0.016390222642156813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,64,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,5120,128,0.009720000127951304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,32,0.005648889061477449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,128,0.012274666792816587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,128,0.014144889182514615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,64,0.012135110795497894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,5120,768,0.011541333463456897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,32,0.012103999654452005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,5120,4096,0.0557182232538859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,65536,0.6665173106723362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,65536,0.2473235660129123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,768,0.02128444446457757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,51200,0.5116284688313802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,5120,3072,0.048727999130884804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,65536,0.751471996307373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,16384,0.17285599973466662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,51200,0.19539911217159697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,12288,0.12328799565633138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,51200,0.5748808648851182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,16384,0.06888533300823636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,16384,0.16539822684393987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,65536,0.765151129828559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,12288,0.05019199848175049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,16384,0.19398844242095947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,12288,0.12681066989898682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,12288,0.14815466933780247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,10240,0.04291911257637871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,10240,0.10824444558885361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,10240,0.12703555160098606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,8192,0.034512887398401894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,7168,0.0802711115943061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,8192,0.08647377623452081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,8192,0.10318666696548462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,7168,0.03129600154028998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,7168,0.07752355602052477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,7168,0.09094666772418553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,6144,0.06672533353169759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,6144,0.02794755498568217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,5120,0.05112266540527344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,10240,0.11012533638212417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,6144,0.0798106657134162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,4096,0.04389866524272495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,5120,0.024614221519894067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,8192,0.0901048911942376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,5120,0.0572213331858317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,3584,0.039854221873813205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,5120,0.0691431098514133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,4096,0.047162665261162647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,4096,0.021470222208235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,3072,0.034471998612085976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,4096,0.057761775122748486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,3584,0.019630221856964957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,3584,0.0425893333223131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,2560,0.029666665527555678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,6144,0.061978664663102895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,3584,0.052095111873414784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,2048,0.023923556009928387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,3072,0.046615110503302686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,2560,0.032976888948016696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,2560,0.04063911239306132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,1536,0.01917155583699544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,2048,0.02755822241306305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,2048,0.03520266546143426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,51200,0.5981591012742784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,1536,0.022456000248591106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,1024,0.014461333552996317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,1536,0.029110223054885864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,1536,0.012847999731699625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,768,0.011702222128709158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,1024,0.011352889239788055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,1024,0.017444445027245414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,1024,0.023002665903833177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,768,0.014855111638704935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,512,0.00906844437122345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,768,0.010560000108347999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,3072,0.017996443642510306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,768,0.0203457772731781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,3072,0.03783199853367276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,512,0.009714666340086196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,2560,0.016283555163277518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,512,0.012423111332787408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,128,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,512,0.01772799922360314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,256,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,2048,0.014668444792429606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,256,0.015152000718646579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,4096,128,0.009319110876984065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,128,0.01073333372672399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,4096,128,0.013040888640615674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,64,0.010699555277824402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,32,0.010682666467295753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,65536,0.45032887988620335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,65536,0.24152443144056532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,256,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,51200,0.35383110576205784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,65536,0.662825796339247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,4096,256,0.011121778024567498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,16384,0.12971999910142687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,65536,0.622803529103597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,32,0.0086666668454806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,51200,0.18826399909125435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,12288,0.10187200042936538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,51200,0.5349502033657497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,16384,0.06548977560467191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,16384,0.14857244491577148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,10240,0.08883466985490586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,16384,0.1598328881793552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,51200,0.4773075315687392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,12288,0.05096888873312208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,12288,0.11586577362484402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,8192,0.07420533233218722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,10240,0.04145866632461548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,12288,0.12386133935716416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,10240,0.09519822067684597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,8192,0.035471111536026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,8192,0.08192888895670573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,10240,0.10581333107418484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,6144,0.05333155393600464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,8192,0.087281776799096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,7168,0.030235555436876085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,7168,0.0690844456354777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,5120,0.0444088876247406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,4096,64,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,6144,0.02773333258099026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,6144,0.06020800272623698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,4096,0.03557866811752319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,6144,0.06853689087761773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,5120,0.024148444334665935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,5120,0.051098667913013034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,3584,0.031543999910354614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,5120,0.05870844258202446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,4096,0.020628444022602506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,4096,0.0425164434644911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,3072,0.027220444546805486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,3584,0.01867911054028405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,4096,0.05001688996950785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,3584,0.038155555725097656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,2560,0.02197422252760993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,3584,0.04490844408671061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,7168,0.059383114178975425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,3072,0.01700711084736718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,3072,0.03408266769515143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,2048,0.0174346665541331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,3072,0.04013777772585551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,2560,0.014502222339312235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,1536,0.014101333088344999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,2560,0.035339554150899254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,2048,0.014256000518798828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,2048,0.02514044443766276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,7168,0.07857244544559054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,2048,0.030393779277801514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,1536,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,1536,0.020433778564135235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,768,0.008763555851247575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,1024,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,1536,0.026317333181699116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,1024,0.01591111057334476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,1024,0.020619556307792664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,768,0.009712888962692684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,768,0.013779555757840475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,256,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,512,0.008976889153321585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,512,0.01720977822939555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,256,0.01054488867521286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,128,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,256,0.008619555168681675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,256,0.014839111102951897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,2560,0.02959377898110284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,64,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3584,128,0.008939555949634975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,128,0.01001422190003925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,32,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,128,0.0118231111102634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,64,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,32,0.010017777482668558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,65536,0.434571557574802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3584,768,0.018583999739752877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3584,512,0.011369778050316704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,65536,0.20518755912780762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,51200,0.3444648848639594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,65536,0.580984910329183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,16384,0.11538133356306289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,51200,0.16251644823286268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,65536,0.6019386715359157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,16384,0.0528639985455407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,16384,0.13800266053941515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,1024,0.01089333328935835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,12288,0.09561510880788167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,51200,0.4731929037306044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,10240,0.08188533120685153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,12288,0.040714667903052434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,16384,0.16017688645256892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3584,512,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,12288,0.1020631127887302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,8192,0.061717331409454346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,10240,0.03584444522857666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,12288,0.12316711743672688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,7168,0.0568444463941786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,8192,0.029183109601338703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,8192,0.0704524450831943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,10240,0.10658133029937744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,6144,0.042908443344963916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,8192,0.08560088608000015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,7168,0.02648266653219859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,7168,0.062485330634646945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,5120,0.03601510988341437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,7168,0.07570844226413302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,6144,0.024010666542583044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,6144,0.055367112159729004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,4096,0.03167555398411221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,5120,0.020960888928837244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,5120,0.04692711101637947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,6144,0.0669999983575609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,51200,0.45691643820868594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,5120,0.05783110857009888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,3584,0.028226666980319556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,4096,0.03984444340070089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,4096,0.01831644442346361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,3072,0.02442577812406752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,4096,0.04909244510862562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,3584,0.01647555496957567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,3584,0.03546844588385688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,3584,0.04404000110096402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,2560,0.019671999745898776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,3072,0.01533777846230401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,10240,0.08678133620156182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,3072,0.03151199883884854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,2048,0.016593777471118502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,2560,0.01400888959566752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,2560,0.027315555347336665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,1536,0.013570666313171387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,2560,0.03461511267556085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,2048,0.02275288932853275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,2048,0.02999911043379042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,1024,0.009720889230569204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,1536,0.011411555939250521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,1536,0.018914666440751817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,1536,0.02500711050298479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,1024,0.010227555202113258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,768,0.007833778030342525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,1024,0.014841778410805596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,1024,0.020251555575264823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,512,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,768,0.009653333160612319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,768,0.012834666503800286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,256,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,512,0.009091555244392818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,512,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,3072,0.039847999811172485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,512,0.01702311138312022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,128,0.004557333472702238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,256,0.008627555436558193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,256,0.009685333404276106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,64,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,2048,0.013046222428480784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,256,0.0141004439857271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,3072,32,0.004280000097221798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,3072,128,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,128,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,128,0.011101333631409539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,64,0.009325332939624786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,3072,32,0.009329777624871995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,65536,0.315558221605089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,65536,0.15849067105187312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,51200,0.24803198708428276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,16384,0.09522310892740886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,65536,0.4887982474433051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,3072,768,0.01835644410716163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,51200,0.38365867402818465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,65536,0.4678719838460286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,12288,0.06013955672581991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,16384,0.04252622193760342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,16384,0.11773778332604302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,10240,0.062168002128601074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,51200,0.37384266323513454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,16384,0.1264666716257731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,12288,0.03301422132386102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,8192,0.04643644558058845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,12288,0.08830489052666558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,10240,0.031028442912631567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,12288,0.09657600190904404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,10240,0.07584089040756226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,7168,0.0406897763411204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,8192,0.02325600054528978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,10240,0.0834862192471822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,8192,0.06242577897177803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,6144,0.037384000089433454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,7168,0.022008889251285132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,8192,0.06897422340181139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,5120,0.030692444907294378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,7168,0.055255111720826894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,6144,0.019334221879641216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,7168,0.06120977799097697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,4096,0.025815111067559984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,6144,0.048001776138941445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,5120,0.017248888810475666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,3584,0.022057778305477563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,5120,0.04086666636996799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,4096,0.015169777803950839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,5120,0.04775999983151754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,51200,0.13053866227467856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,3072,0.01924000018172794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,4096,0.039903998374938965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,3584,0.01407999959256914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,2560,0.01704266667366028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,3584,0.031077331966824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,3584,0.036044445302751325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,3072,0.012785777449607849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,3072,0.027295110954178706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,2048,0.014115555418862237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,3072,0.03271822134653727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,2560,0.011922666596041786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,2560,0.024075556132528517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,1536,0.01146666705608368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,2560,0.02914311157332526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,2048,0.010815111299355825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,2048,0.01998844411638048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,1024,0.008619555168681675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,2048,0.025626666016048853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,1536,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,1536,0.016873778568373788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,768,0.007268444531493717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,1536,0.02104355560408698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,1024,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,1024,0.013341333303186627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,512,0.005872000008821487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,1024,0.017512889371977914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,768,0.008925333619117737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,768,0.011642666326628791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,4096,0.03448266784350077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,768,0.016174222032229107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,512,0.009655111365848118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,512,0.014249778456158109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,256,0.00793333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,128,0.003839111162556542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,256,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,256,0.01276177747382058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,64,0.003543111185232798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,128,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,32,0.0037671112351947357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,128,0.008637333081828224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,128,0.010675555302037133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,64,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2560,32,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,65536,0.293175114525689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2560,6144,0.05458577805095249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,51200,0.23028887642754448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,65536,0.15284266736772326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2560,256,0.004592888885074192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2560,512,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,16384,0.08619378010431926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,51200,0.12079555458492702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,65536,0.4570924441019694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,12288,0.06578310992982653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,51200,0.3380720085567898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,16384,0.04076177875200907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,16384,0.10301066769493951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,10240,0.05669689178466797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,16384,0.12437689304351807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,12288,0.03206933206982083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,51200,0.3620897663964166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,12288,0.07890222469965617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,10240,0.028767112228605483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,12288,0.09493333101272583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,10240,0.06724711259206136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,7168,0.03863999909824795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,10240,0.08049155606163873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,8192,0.05487111210823059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,8192,0.022967111733224656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,6144,0.03409333361519708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,8192,0.06705511278576322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,7168,0.021039111746682063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,7168,0.049229333798090615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,5120,0.028564443190892536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,65536,0.4332328902350531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,7168,0.059895111454857715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,6144,0.019122666782803006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,6144,0.04263733492957222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,4096,0.024008888337347243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,6144,0.05350755651791891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,5120,0.016908443636364408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,5120,0.036884443627463445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,3584,0.021477333373493616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,4096,0.0147706667582194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,5120,0.04598311252064175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,4096,0.031349334451887347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,3584,0.013799111048380533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,4096,0.03932177689340379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,2560,0.016599999533759225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,3072,0.012703110774358114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,3584,0.035791112316979304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,3072,0.02465244465404087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,8192,0.04380177789264255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,2048,0.013941332697868347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,3072,0.03135111265712314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,2560,0.011780444118711682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,2560,0.021320000290870667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,1536,0.010882666541470421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,2048,0.010553778045707278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,2560,0.02758311066362593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,2048,0.018246221873495314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,1024,0.008271111382378472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,2048,0.024494222468800012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,1536,0.0097884444726838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,1536,0.015180443723996481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,768,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,1536,0.02086755633354187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,1024,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,512,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,1024,0.017303999927308824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,768,0.00870666652917862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,768,0.010442666709423065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,768,0.015885333220163982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,3072,0.019464888506465487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,512,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,512,0.00903555585278405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,3584,0.027979556057188246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,512,0.01443288889196184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,128,0.003912000192536248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,256,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,256,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,64,0.003875555677546395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,256,0.012799111505349478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,2048,128,0.007810666329330868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,32,0.0042239998777707415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,128,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,2048,128,0.01036711119943195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,64,0.007818666597207388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,65536,0.24962933858235678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,32,0.007923555870850882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,51200,0.1955582168367174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,65536,0.14520800113677979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,65536,0.33829689025878906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,16384,0.06531733274459839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,51200,0.26315376493665904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,2048,256,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,51200,0.11496178309122722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,65536,0.4570693439907498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,12288,0.0472924444410536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,16384,0.03887822230656942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,16384,0.08674310975604588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,10240,0.04074666566318936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,51200,0.38215822643703884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,16384,0.12331467204623753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,12288,0.030799110730489094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,12288,0.06458222203784518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,8192,0.03157066636615329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,12288,0.093213333023919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,10240,0.027393778165181477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,10240,0.05510044429037306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,8192,0.021736888421906367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,8192,0.04517689016130236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,10240,0.08198489083184136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,6144,0.02296888828277588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,8192,0.06655377811855741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,7168,0.020025778147909377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,2048,1024,0.012106666962305704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,7168,0.04053333401679993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,5120,0.019430221782790292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,7168,0.0594044460190667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,6144,0.017954667409261067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,6144,0.03568533394071791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,4096,0.016872889465755887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,6144,0.05262310968505012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,5120,0.01610577768749661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,5120,0.0304097765021854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,3584,0.015301333533393012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,5120,0.04526933365397983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,4096,0.014478221535682678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,4096,0.0255031105544832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,4096,0.038429333104027644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,3584,0.013399110900031196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,3584,0.022861333356963262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,3584,0.0337119996547699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,3072,0.012435555458068848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,3072,0.020432889461517334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,2048,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,7168,0.02773333258099026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,2560,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,2560,0.017910222212473553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,1536,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,2560,0.02715644405947791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,2048,0.010388444695207808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,2048,0.01549600064754486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,1024,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,2048,0.023969777756267126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,1536,0.013066666821638743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,1536,0.020604444874657523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,1024,0.008775110873911116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,768,0.00544088871942626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,1024,0.010690666735172272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,1024,0.016861332787407767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,768,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,3072,0.013203555511103736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,768,0.009182222187519073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,2560,0.011510222322411008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,256,0.003568888952334722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,512,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,512,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,512,0.014008000493049622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,3072,0.031096001466115315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,128,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,256,0.007880888879299164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,256,0.007418666448858049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,64,0.003148444410827425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,256,0.012720888687504662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,32,0.0031635556370019913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,128,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,128,0.0069439998931354946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,128,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1536,1536,0.009728000395827824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,64,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,65536,0.17766222688886854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1536,32,0.007251555720965068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,51200,0.1450186702940199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,65536,0.09986578093634711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,65536,0.27978844112820095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1536,512,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,16384,0.04557955596182081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,51200,0.07978933387332492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1536,768,0.015475556254386902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,65536,0.3152302106221517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,51200,0.22237687640719941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,12288,0.0347671111424764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,16384,0.027286221583684284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,16384,0.07245066430833605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,10240,0.029853334029515583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,51200,0.2500986788007948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,12288,0.022036444809701707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,16384,0.08772799703809951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,12288,0.05513244536187914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,10240,0.02036533256371816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,12288,0.06739999850591023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,7168,0.0211208893193139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,10240,0.0465280016263326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,8192,0.016193777322769165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,8192,0.0388373335202535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,6144,0.018383999665578205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,8192,0.04819733235571119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,7168,0.014771555860837301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,7168,0.0343466666009691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,5120,0.016093333562215168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,7168,0.04378844300905863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,6144,0.01348888874053955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,6144,0.030615998639000788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,4096,0.013577777478430005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,6144,0.0388008885913425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,5120,0.012352888782819113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,5120,0.026313778426912095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,3584,0.012445333103338877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,5120,0.03379377722740173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,4096,0.01108444482088089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,4096,0.022037333912319605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,3072,0.010675555302037133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,4096,0.028179556131362915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,3584,0.010450666977299584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,3584,0.0199608885579639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,8192,0.023783110909991797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,2560,0.009379555781682333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,3584,0.025880000657505457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,3072,0.010019555687904358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,2048,0.007821333077218797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,3072,0.01791555517249637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,2560,0.009353777600659264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,10240,0.05762400229771932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,2560,0.015853333804342482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,3072,0.023584889041052923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,2048,0.00903911143541336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,2560,0.020967110991477966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,1024,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,2048,0.013783111340469785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,1536,0.008374222450786168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,2048,0.018704889549149405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,768,0.0046640001237392426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,1536,0.011995555626021491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,1024,0.007968000239796108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,1024,0.009516444471147325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,512,0.003922666526503033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,768,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,1024,0.014093332820468478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,768,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,256,0.003216000066863166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,768,0.012924444344308643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,512,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,128,0.00314666661951277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,512,0.007249777515729268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,512,0.012120000190205045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,256,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,256,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,64,0.0028479999552170434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,256,0.011358222199810876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,1024,128,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,128,0.006311999840868845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,128,0.008639111287064022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,64,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,65536,0.14504978391859266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,1024,32,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,1536,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,51200,0.11497067080603705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,65536,0.09849422507815891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,65536,0.23938311470879448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,16384,0.03838933176464505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,51200,0.07922844754325019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,1024,1536,0.016550223032633465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,51200,0.18865066104465059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,12288,0.02793244520823161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,65536,0.31129688686794704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,16384,0.026650667190551758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,16384,0.06226666768391927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,51200,0.24279912312825522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,10240,0.02299555473857456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,12288,0.021700445148679946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,16384,0.08628088898128933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,12288,0.04775377776887682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,8192,0.01867022282547421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,10240,0.01980977753798167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,12288,0.06698399782180786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,10240,0.04110488957828946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,7168,0.017059556312031217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,10240,0.057499554422166615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,8192,0.01593155496650272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,6144,0.015223110715548197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,8192,0.033831109603246055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,8192,0.048488888475630015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,7168,0.014521777629852295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,7168,0.03064799971050686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,5120,0.01333688861793942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,7168,0.04275111026234097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,6144,0.013413333230548434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,6144,0.027090667022599116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,4096,0.011156444748242697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,6144,0.03857600026660495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,5120,0.012134222520722283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,3584,0.009953777823183272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,5120,0.023018666439586218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,5120,0.03251911203066508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,4096,0.011026666396194033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,4096,0.019612444771660697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,3072,0.008991999758614434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,4096,0.02797599964671665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,3584,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,3584,0.017882666654056974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,2560,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,3584,0.02567555507024129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,3072,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,2048,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,3072,0.01611111064751943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,3072,0.02345422241422865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,2560,0.009544889132181803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,1024,32,0.0028524444335036804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,2560,0.014320888453059726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,1536,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,2560,0.020630222227838304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,2048,0.009016888837019602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,1024,0.0051893335249688895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,2048,0.012431110772821637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,1536,0.008632000121805403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,2048,0.018910222583346896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,768,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,1536,0.011058666639857821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,1024,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,1536,0.01612711118327247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,512,0.0038408889538711975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,1024,0.008905777500735389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,768,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,1024,0.013585777746306526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,256,0.0032035555276605818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,512,0.007325333025720384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,768,0.01277511070171992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,512,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,128,0.0028239999794297745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,512,0.01143022212717268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,256,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,64,0.0025840000145965153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,256,0.006223111102978389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,768,32,0.0028320000403457214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,768,128,0.006909333169460297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,128,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,65536,0.10224888722101848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,128,0.008617777791288164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,64,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,32,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,51200,0.07700088951322767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,65536,0.09785866737365723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,16384,0.03156800071398417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,65536,0.20080178313785127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,51200,0.07890933089786105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,51200,0.1598337756262885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,12288,0.022672888305452134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,65536,0.2986275619930691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,16384,0.02522755497031742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,16384,0.052816000249650746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,51200,0.23567732175191244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,10240,0.01977422171168857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,16384,0.0829004446665446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,768,768,0.007689778175618913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,12288,0.02088266611099243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,8192,0.016728889611032274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,12288,0.04092622134420607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,10240,0.019274666905403137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,12288,0.06393777661853366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,7168,0.014779556128713818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,10240,0.03513155712021722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,8192,0.015447111593352424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,10240,0.05520711011356778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,768,256,0.01053600013256073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,8192,0.029088000456492107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,7168,0.013985777894655863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,8192,0.04572977622350057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,5120,0.011023110813564725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,7168,0.026340444882710774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,6144,0.012881778180599213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,7168,0.040249778164757624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,4096,0.009391999906963771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,5120,0.011744000017642975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,6144,0.036165333456463285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,5120,0.02032888929049174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,3584,0.008646222452322641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,5120,0.031005332867304485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,4096,0.010665777656767102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,4096,0.017163554827372234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,3072,0.007780444290902879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,4096,0.026641777820057336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,3584,0.010260444548394945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,3584,0.015615999698638916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,2560,0.0069439998931354946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,3584,0.024495111571417913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,3072,0.009682666924264696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,3072,0.014112000664075216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,2048,0.006217777729034424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,2560,0.00923911150958803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,2560,0.012685333689053854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,6144,0.01308444473478529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,2560,0.019923556182119582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,1536,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,2048,0.009038222332795462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,2048,0.011143111520343356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,1024,0.004229333251714706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,2048,0.018209778600268893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,1536,0.008346666892369589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,6144,0.023258666197458904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,1536,0.00960711141427358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,768,0.0038319999972979226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,1536,0.016000888413853116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,1024,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,512,0.0034355556385384668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,1024,0.013426666458447775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,768,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,768,0.007083555890454187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,256,0.0028693332440323303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,768,0.012621333201726278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,512,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,512,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,128,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,512,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,256,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,64,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,256,0.005967111223273807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,512,32,0.002597333242495855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,256,0.010737777584128909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,128,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,3072,0.02241777711444431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,128,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,65536,0.05875644418928358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,512,128,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,51200,0.04742133286264208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,64,0.005628444254398346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,512,32,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,65536,0.09745777977837457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,16384,0.01755555636352963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,65536,0.1571306652492947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,51200,0.07853955692715116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,51200,0.12399110529157852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,12288,0.01692800058258904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,16384,0.02442399991883172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,512,1024,0.008006222546100616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,16384,0.04206577936808268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,51200,0.2334968778822157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,10240,0.015222221612930298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,16384,0.08135110802120633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,12288,0.02033955521053738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,8192,0.012728888955381183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,12288,0.03278133273124695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,10240,0.01924533314175076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,12288,0.06345511145061918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,7168,0.011438222395049201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,10240,0.02849777870708042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,8192,0.015241778559154935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,10240,0.054471111959881253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,6144,0.00998044427898195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,8192,0.023411555422676936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,7168,0.014140443669425117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,8192,0.04496888981925117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,7168,0.021143111917707656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,5120,0.008622222476535374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,6144,0.012786666552225748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,7168,0.03969600134425693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,4096,0.007252444823582967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,6144,0.018917333748605516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,5120,0.011762667033407422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,6144,0.035232888327704534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,3584,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,5120,0.0166640000210868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,4096,0.010665777656767102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,5120,0.030528889762030706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,3072,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,4096,0.014488000008794995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,3584,0.010049777726332346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,4096,0.026329777306980554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,3584,0.013431111143694984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,65536,0.2932444413503011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,3072,0.009684444301658207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,3584,0.024343111448817786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,2048,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,3072,0.012048000262843238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,3072,0.021976888179779053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,2560,0.01089511149459415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,1536,0.004182222402758068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,2560,0.01975200076897939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,2048,0.008830221990744272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,2048,0.009633777870072259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,1024,0.0032266666077905228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,2048,0.01755911111831665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,1536,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,1536,0.008350222474998897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,768,0.003160888950030009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,1536,0.015489778584904142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,1024,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,1024,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,512,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,1024,0.013088000317414602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,768,0.0075866662793689305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,768,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,256,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,768,0.012113778127564324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,512,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,512,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,128,0.0021840000732077495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,2560,0.005483555710977978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,512,0.011027555498811932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,256,0.006907555378145642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,256,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,64,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,256,32,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,128,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,256,2560,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,128,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,65536,0.04420888755056593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,128,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,64,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,51200,0.035192890299691096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,256,32,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,65536,0.09734133217069839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,16384,0.011055111057228513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,65536,0.15663111209869385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,51200,0.07782755295435588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,51200,0.1230284505420261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,12288,0.009516444471147325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,16384,0.024109333753585815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,16384,0.04160977734459771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,51200,0.231948455174764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,10240,0.01055022246307797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,16384,0.08065066734949748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,12288,0.020336889558368258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,12288,0.032398223876953125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,8192,0.009641778137948778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,12288,0.06264711088604398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,10240,0.01888177792231242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,10240,0.02780533168050978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,7168,0.010675555302037133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,256,256,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,8192,0.01516266663869222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,8192,0.023008000519540574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,6144,0.009381333159075843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,7168,0.01407911048995124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,8192,0.043904887305365674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,7168,0.020959110723601446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,5120,0.008254222571849823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,7168,0.03907555672857497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,6144,0.012823999755912356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,4096,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,5120,0.011720889144473605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,6144,0.035306665632459856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,5120,0.01648977730009291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,65536,0.2928168773651123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,3584,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,5120,0.030513776673210993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,4096,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,4096,0.014470222923490735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,3072,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,3584,0.010055111514197456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,4096,0.026050666968027752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,2560,0.005076444397370021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,3584,0.02400355537732442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,3072,0.009673777553770278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,3072,0.02196888956758711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,2048,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,10240,0.05247999893294441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,2560,0.010967111421955956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,2560,0.019332443674405415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,2048,0.00868088917599784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,1536,0.003855111284388436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,2048,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,2048,0.017567111386193168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,1024,0.0032115555885765287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,6144,0.0188071115149392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,1536,0.008303110798199972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,1536,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,768,0.0028239999794297745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,1536,0.015450666348139444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,1024,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,1024,0.007342221836249034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,1024,0.013071111506885953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,768,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,512,0.0025315555847353404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,768,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,768,0.012056888805495368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,256,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,512,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,512,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,512,0.010676444404655032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,2560,0.009317333499590555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,128,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,256,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,256,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,64,0.0018933332628673976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,256,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,128,32,0.001959111127588484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1792,128,128,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,128,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,65536,0.042938666211234204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,51200,0.033336000310050115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,64,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,16384,0.013581333061059317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,32,0.004998222407367495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,12288,0.011372444530328115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,65536,0.15623288684421116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,10240,0.010311110979980892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,51200,0.12283110618591309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,16384,0.041491554843054876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,8192,0.00942133367061615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,12288,0.03196977906756931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,7168,0.008857777549160851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,3584,0.01333688861793942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,6144,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,5120,0.007268444531493717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,128,3072,0.012070222033394707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,10240,0.027499554885758296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,4096,0.006682666639486949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,8192,0.023004444109068975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,7168,0.020937777227825586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,6144,0.018650665879249573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,5120,0.016549333930015564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,2560,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,4096,0.014371555712487964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,2048,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,3584,0.01309866706530253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,1536,0.0038702223036024305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,3072,0.01204711116022534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,1024,0.0032071111102898917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,2560,0.010753778119881948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,768,0.002879111096262932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,1536,0.00849422232972251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,512,0.002528888897763358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,1024,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,256,0.0022942223068740633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,768,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,128,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,64,0.001999999913904402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,512,0.005977777971161737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,32,0.0021359999146726397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,256,0.005326222214433882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,128,0.005170666509204441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,3584,0.00620977787507905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,65536,0.04439377784729004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,64,0.005065777649482091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,51200,0.03395911057790121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,64,3072,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,16384,0.009208888643317753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,32,0.004960888789759742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,12288,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,10240,0.007000889215204451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,8192,0.006551110910044775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,51200,0.12264266279008652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,65536,0.1559342278374566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,12288,0.031883554326163396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,7168,0.006004444426960415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,10240,0.02747022277779049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,6144,0.009287999735938178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,5120,0.008580444587601556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,8192,0.023050665855407715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,4096,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,64,2048,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,3584,0.006990222467316522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,3072,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,7168,0.02091911103990343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,6144,0.01865511139233907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,2560,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,5120,0.016531555189026725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,2048,0.005114666703674528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,4096,0.014309333430396186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,1536,0.003664888855483797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,3584,0.013112000293201871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,1024,0.003168888803985384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,3072,0.012063999970753988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,2560,0.011020444333553314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,2048,0.00962755580743154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,512,0.0024906665914588505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,1536,0.008424888882372115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,1024,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,128,0.0019040000107553268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,768,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,512,0.005802666561471091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,256,0.005377777748637729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,64,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,128,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,32,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,16384,0.041561779048707746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,64,0.004966222163703707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1792,32,32,0.005107555538415909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,768,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1792,32,256,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,16384,0.6998106638590494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,16384,2.0022506713867188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,16384,2.2327262030707464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,12288,0.5157039960225424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,51200,2.4231698777940536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,12288,1.6396178139580622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,12288,1.584366268581814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,10240,1.424584918551975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,10240,0.4366604487101237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,10240,1.3573377397325304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,8192,1.0949689017401802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,8192,0.3495466709136963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,8192,1.0304027133517795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,7168,0.3423777686225043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,7168,1.1691395441691081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,7168,0.8818871180216471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,6144,0.931971549987793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,6144,0.750707573360867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,6144,0.2753324508666992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,51200,6.613321940104167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,5120,0.2287688785129123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,4096,0.5132248666551378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,5120,0.851085344950358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,5120,0.6241502232021755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,4096,0.5209137598673502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,3584,0.5406568845113119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,3584,0.17012266318003336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,4096,0.19701777564154732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,3584,0.4501342243618435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,3072,0.17624355687035453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,3072,0.3784204324086507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,2560,0.3809795644548204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,2560,0.13395733303493923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,2048,0.30452267328898114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,2048,0.117648892932468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,1536,0.22125866678025988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,51200,8.200162251790365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,2048,0.2625813219282362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,1536,0.09323733382754856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,1024,0.16227289040883383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,1536,0.23050665855407715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,768,0.12460445033179389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,3072,0.38787023226420086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,1024,0.16626311673058405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,768,0.06653066476186116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,256,0.055905779202779136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,512,0.05913155608707004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,512,0.10905155870649551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,128,0.04160977734459771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,256,0.05391644438107809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,2560,0.3241911199357775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,256,0.09768444299697876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,128,0.05134400063090854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,128,0.09129244751400417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,64,0.09124888976414998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,65536,1024,0.07198755608664618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,32,0.0920613341861301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,512,0.07933688826031156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,65536,768,0.12877066930135092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,64,0.0398888885974884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,65536,32,0.04318755533960131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,16384,1.4250382317437067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,51200,1.9335447947184246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,65536,2.425906711154514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,16384,1.7719395955403645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,16384,0.5338346693250868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,51200,4.597201029459636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,12288,1.3894773059421117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,12288,0.4042293230692546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,12288,1.2440310584174263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,65536,6.1026357014973955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,10240,0.34137866232130265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,10240,1.3815511067708333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,8192,0.28159112400478786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,8192,0.74053865008884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,10240,1.0787102381388347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,51200,6.407408820258246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,7168,0.907448026869032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,7168,0.24646578894721138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,8192,0.8317137824164497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,6144,0.6596933470831977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,7168,0.742012447781033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,6144,0.601576911078559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,5120,0.4842764536539714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,5120,0.20711555745866564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1792,128,128,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,4096,0.47070932388305664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,4096,0.15220355987548828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,4096,0.39320354991488987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,3584,0.16153066688113743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,3072,0.31701956854926217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,3584,0.34595733218722874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,65536,8.820995754665798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,2560,0.23323376973470053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,3072,0.2944168779585096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,3072,0.11999555428822835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,6144,0.21189867125617134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,2560,0.10689777798122829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,2560,0.2706480026245117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,5120,0.4965253406100803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,1536,0.15924533208211264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,2048,0.08793689144982232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,1024,0.11660622225867377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,1536,0.07059911224577162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,3584,0.3853448761834039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,2048,0.21474933624267578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,1536,0.16369688510894775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,768,0.09825066725413005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,512,0.0719004472096761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,1024,0.05732622411515978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,768,0.05245511068238152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,256,0.04711733261744181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,768,0.09918933444552952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,512,0.047592888275782265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,512,0.0862026678191291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,128,0.03602755400869582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,256,0.044064889351526894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,64,0.03115644388728672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,256,0.07733511262469821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,32,0.03230399886767069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,51200,128,0.04183822207980686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,128,0.07280621925989787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,51200,2048,0.18786666128370497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,64,0.07275466786490546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,32,0.07330666648017035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,51200,1024,0.12426400184631348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,16384,0.5450497733222114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,51200,0.5986506674024794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,65536,0.7780293358696831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,16384,0.2125991053051419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,16384,0.5376702414618598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,51200,1.5802222357855902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,12288,0.4015759891933865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,12288,0.14588977230919734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,10240,0.3529493278927273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,65536,2.1494223276774087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,10240,0.11954489019181992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,8192,0.28285956382751465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,12288,0.4754008717007107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,51200,1.9131084018283422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,7168,0.24219022856818306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,10240,0.3577875561184353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,8192,0.10815999905268352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,7168,0.0882986651526557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,6144,0.1725564400355021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,8192,0.29927200741238064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,7168,0.24578489197625053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,6144,0.07459200090832181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,5120,0.18215555614895293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,65536,2.662815941704644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,4096,0.15384444925520155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,5120,0.0646231108241611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,6144,0.22924089431762695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,3584,0.1402017805311415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,4096,0.05396622088220385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,5120,0.16566133499145508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,3072,0.10738311211268108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,3584,0.0542791121535831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,2560,0.09810933139589097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,4096,0.13731822702619764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,3072,0.0484186675813463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,3584,0.11956355306837295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,2048,0.08049866888258192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,1536,0.05725510915120443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,3072,0.10537866751352946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,2560,0.03830666674507989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,2048,0.03311199943224589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,2560,0.0877786676088969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,768,0.028171555863486394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,1536,0.028734223710166082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,512,0.02292888859907786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,1024,0.022135110365019903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,768,0.020672889219390023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,1024,0.044116444057888456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,768,0.03593688872125413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,256,0.015944888194402058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,512,0.018637332651350234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,128,0.011972444752852121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,512,0.029994666576385498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,256,0.016755556066830952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,64,0.011040888726711273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,16384,128,0.016729777057965595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,256,0.02622577713595496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,32,0.010703999963071613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,128,0.025398222936524287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,64,0.025405334101782903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,32,0.025091555383470323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,16384,1024,0.03666933377583822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,1536,0.0580693350897895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,16384,2048,0.07312977976269193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,65536,0.5939466688368056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,16384,0.3571751117706299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,51200,0.4610568682352702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,65536,1.543183962504069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,51200,1.3298578262329102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,12288,0.3601253297593858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,16384,0.4990968704223633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,12288,0.10946044656965469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,51200,1.6305466757880316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,10240,0.2820826636420356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,65536,2.0445885128445096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,12288,0.3093973265753852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,10240,0.09432088666492039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,10240,0.25294576750861275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,8192,0.2199617756737603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,8192,0.07792088720533583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,7168,0.16549244191911486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,6144,0.16695200072394478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,7168,0.0675271087222629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,8192,0.278714656829834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,6144,0.05865155325995552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,7168,0.17729600270589194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,5120,0.14254666699303523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,4096,0.09351022375954522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,6144,0.15449244446224636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,5120,0.05199644300672743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,16384,0.13974400361378989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,3584,0.07953955729802449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,4096,0.052593777577082314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,5120,0.1284373336368137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,3072,0.07278755638334486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,2560,0.06228444311353895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,4096,0.10236444738176133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,3072,0.03834933373663161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,2048,0.058690667152404785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,3584,0.09094311131371392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,2560,0.03038755390379164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,1536,0.03757955630620321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,2560,0.06866666343477038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,2048,0.0262106673585044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,1024,0.026085333691702947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,768,0.022052443689770166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,1536,0.021801778011851843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,1536,0.045202665858798556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,512,0.01685333251953125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,768,0.01687466601530711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,768,0.02864355511135525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,256,0.012734221915404001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,512,0.015431111057599386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,512,0.023592000206311543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,128,0.009415999882751042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,256,0.014111111561457316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,256,0.021037333541446265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,64,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,128,0.01422755585776435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,3584,0.03865866528617011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,12288,32,0.008482666479216682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,128,0.020280000236299302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,3072,0.08014400137795342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,64,0.020290666156344943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,32,0.020226667324701946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,2048,0.05661333269543118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,12288,1024,0.018537777993414137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,12288,1024,0.0347324444188012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,16384,0.3215999868181017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,65536,0.48712889353434247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,51200,0.3904426627688938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,51200,0.8804684215121799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,16384,0.3331848780314128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,65536,1.2678409152560763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,16384,0.1184426678551568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,12288,0.244142214457194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,10240,0.2132248878479004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,10240,0.07801777786678739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,8192,0.16671466827392578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,12288,0.3445617887708876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,7168,0.14609866672092015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,8192,0.06327911218007405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,65536,1.5690391328599718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,8192,0.1817022164662679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,6144,0.1300897730721368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,7168,0.06507733133104113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,5120,0.10360266764958699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,6144,0.05050488975312975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,7168,0.19768799675835502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,6144,0.1322480042775472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,4096,0.08875199821260239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,5120,0.04253333475854662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,3584,0.07692444324493408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,4096,0.03610310951868693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,5120,0.11622222264607747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,3072,0.05644622113969591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,4096,0.08924266364839341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,3584,0.03226044442918565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,2560,0.0528639985455407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,3584,0.0784097777472602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,3072,0.029323554701275293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,2048,0.04567644331190321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,12288,0.1004142231411404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,3072,0.06831200255288018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,1536,0.03355377912521362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,2560,0.026749332745869953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,2048,0.022924444741672937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,2560,0.058676441510518394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,1024,0.02453688946035173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,51200,1.1678827073838975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,1536,0.019519999623298645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,2048,0.04923644330766466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,768,0.01903377804491255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,512,0.01385155568520228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,1536,0.039467556609047785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,1024,0.01647111111217075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,256,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,768,0.015083556373914083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,512,0.013591999808947245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,768,0.024991111622916326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,128,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,512,0.020620443754725985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,256,0.012720000412729053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,64,0.007429333196745978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,256,0.01825777855184343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,10240,128,0.012893333203262754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,10240,32,0.007376000285148621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,128,0.017644445101420086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,64,0.017736888594097562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,32,0.01755288905567593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,10240,0.23534843656751844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,65536,0.41781245337592227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,10240,1024,0.030243555704752605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,51200,0.7005013359917535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,16384,0.26324354277716744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,16384,0.10932799842622544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,51200,0.3193688922458225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,12288,0.2192924420038859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,65536,1.2111893759833443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,51200,0.9420382181803385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,16384,0.27031556765238446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,12288,0.07659644550747342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,10240,0.18453066878848604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,8192,0.1441333293914795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,10240,0.06716977887683444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,12288,0.22759111722310385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,10240,0.17259732882181802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,8192,0.06984355714586046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,6144,0.10804000165727402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,7168,0.0480328897635142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,7168,0.12453777260250515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,5120,0.08084266715579562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,6144,0.04159644577238295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,4096,0.06859999895095825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,65536,0.9676533804999458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,6144,0.12222133742438422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,5120,0.03615644574165344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,3584,0.06924533181720309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,5120,0.0902151134279039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,4096,0.031054221921496924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,3072,0.057414220439063184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,3584,0.028375999795065984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,2560,0.047192000680499606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,3584,0.06626222531000774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,2048,0.035924444595972695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,3072,0.025408888856569927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,2560,0.023112000690566167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,1536,0.02717777755525377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,2560,0.049026668071746826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,2048,0.01982399986849891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,1024,0.019800000720553927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,1536,0.016930666234758165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,2048,0.040750222073660955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,7168,0.12213333447774251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,768,0.016216889023780823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,1536,0.03326311045222812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,512,0.01240355521440506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,1024,0.014264000786675347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,768,0.013134222063753339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,256,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,768,0.021329777108298406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,512,0.012196444802814059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,128,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,512,0.017496888836224873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,256,0.011360888679822287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,64,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,256,0.015776000089115567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,8192,128,0.011408000356621213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,8192,32,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,128,0.015117333994971381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,64,0.015134221977657743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,32,0.015120888749758402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,8192,0.13801333639356825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,1024,0.025632888078689575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,65536,0.37393866644965273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,65536,0.947397338019477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,51200,0.6907413270738391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,16384,0.24031289418538412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,4096,0.07252711057662964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,65536,1.0729138056437175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,16384,0.0921742253833347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,12288,0.1800008879767524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,16384,0.27404356002807617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,12288,0.0691937804222107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,10240,0.16381422678629556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,12288,0.18073599868350557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,8192,0.12940888934665257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,10240,0.06033600038952298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,51200,0.8268053266737195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,10240,0.17670755916171602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,7168,0.11649955643547906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,8192,0.04922222097714742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,6144,0.10005866818957859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,7168,0.04373688830269707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,8192,0.12251199616326226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,7168,0.11574133237202962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,5120,0.085789336098565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,6144,0.03827644387880961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,4096,0.06513422065311007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,6144,0.09540977742936875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,5120,0.034662223524517484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,3584,0.06082044707404243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,5120,0.07969244321187337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,4096,0.065010666847229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,3584,0.025755556093321905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,3584,0.05840266413158841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,8192,3072,0.05611733595530192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,2560,0.03789155681927999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,3072,0.023429332507981196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,2048,0.0328435566690233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,3072,0.049828443262312144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,2560,0.021534222695562575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,1536,0.02292088833120134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,2560,0.043549332353803843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,2048,0.018657777044508193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,1024,0.016640888320075143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,2048,0.03711377912097507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,1536,0.015962666935390897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,768,0.013439111411571503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,1536,0.030047999487982854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,1024,0.01348355578051673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,512,0.011113777756690979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,1024,0.023099554909600153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,768,0.01276177747382058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,256,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,768,0.019211555520693462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,512,0.011733333269755045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,4096,0.028439998626708984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,128,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,512,0.0158595558669832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,3072,0.04695644312434726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,64,0.005836444596449534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,51200,0.2902924484676785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,256,0.010990222295125326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,7168,32,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,256,0.014456000592973499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,128,0.013767999907334646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,64,0.013816888961527081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,7168,32,0.013809777796268463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,16384,0.1831937763426039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,65536,0.33744178877936465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,51200,0.5984284612867568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,51200,0.26332799593607586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,16384,0.0787715580728319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,12288,0.16382488939497206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,16384,0.21601511372460258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,51200,0.7188880178663465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,65536,0.9329048792521158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,10240,0.13850311438242593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,12288,0.062223109934065074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,12288,0.16679733329349092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,8192,0.09725066688325669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,10240,0.05853955613242256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,7168,0.0900542206234402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,8192,0.04323377874162462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,6144,0.07888266775343153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,8192,0.13355555799272326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,7168,0.04266400138537089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,5120,0.06669333246019152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,6144,0.03456711106830173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,7168,0.09724266661538018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,4096,0.0514613323741489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,6144,0.08332533306545682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,5120,0.030431111653645832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,5120,0.07278577486673991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,4096,0.0255404445860121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,65536,0.679166211022271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,4096,0.05782844622929891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,3072,0.03797422183884515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,2560,0.03246666656600104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,3584,0.022913777165942725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,3072,0.021009777983029682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,3584,0.051791111628214516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,2048,0.02572800053490533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,2560,0.01923377811908722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,3072,0.04496888981925117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,1536,0.020576889316240948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,2048,0.016735111673672993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,1024,0.015226667126019796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,2048,0.03333688775698344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,7168,128,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,1536,0.01442577772670322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,10240,0.1340213351779514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,768,0.012707555459605323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,1536,0.026949332820044622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,1024,0.012336889074908363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,512,0.009490666290124258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,1024,0.020562666985723708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,256,0.007240888973077138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,768,0.017533333765135873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,512,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,128,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,512,0.014468444718254937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,256,0.01000711073478063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,64,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,256,0.013082666529549493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,128,0.010022222167915767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,32,0.0058222222659322954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,128,0.012759110993809171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,64,0.0127368892232577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,32,0.01275288893116845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,6144,2560,0.039600888888041176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,65536,0.519760873582628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,65536,0.269707547293769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,51200,0.4847635693020291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,6144,768,0.011691555380821228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,16384,0.1631244421005249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,51200,0.20825866858164468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,65536,0.7709457609388563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,12288,0.12603999508751765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,16384,0.06509155697292753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,6144,3584,0.044827553961012095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,51200,0.6074995464748806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,10240,0.11531378163231744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,12288,0.051297777228885226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,16384,0.2008862230512831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,8192,0.0884008871184455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,7168,0.07450489203135173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,12288,0.15620356135898167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,10240,0.11484355396694607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,8192,0.03567555546760559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,6144,0.05645155244403415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,8192,0.09845866759618123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,7168,0.03418488966094123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,5120,0.0457822216881646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,6144,0.028845333390765723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,7168,0.08454577790366279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,4096,0.03787111242612203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,6144,0.07198577457004122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,5120,0.02568000058333079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,3584,0.03341866532961527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,3072,0.02898755669593811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,5120,0.060672885841793485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,3584,0.020651555723614164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,4096,0.04950222373008728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,3072,0.0184542222155465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,3584,0.0451564457681444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,2048,0.021648888786633808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,3072,0.03985866573121812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,1536,0.016729777057965595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,2560,0.034761776526769005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,2048,0.014750222365061441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,2048,0.02897955642806159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,1536,0.013122667041089801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,10240,0.04439022143681844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,1536,0.02336799932850732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,768,0.010635555618339114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,1024,0.011400000088744693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,512,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,1024,0.018232888645595975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,768,0.01070844464831882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,256,0.006228444476922353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,512,0.009993777506881291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,512,0.013077333569526672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,128,0.005698666804366642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,256,0.009378666679064432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,256,0.011787555283970304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,2560,0.02823288904296027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,64,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,32,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,128,0.009463999834325578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,128,0.011394666300879585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,2560,0.01667288939158122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,64,0.011357333097192975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,32,0.011432888607184092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,5120,1024,0.012710221939616732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,65536,0.49093421300252277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,65536,0.24193867047627768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,51200,0.4210044542948405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,16384,0.13679556051890054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,51200,0.19150043858422172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,5120,768,0.015267555912335714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,12288,0.10001244147618611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,65536,0.657775984870063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,16384,0.07364888985951741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,10240,0.1031351089477539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,51200,0.5049653583102757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,5120,4096,0.022247999906539917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,12288,0.047356443272696606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,8192,0.07943110995822482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,12288,0.11286311679416233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,10240,0.05150577757093641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,7168,0.0644435551431444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,6144,0.05595022439956665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,8192,0.03340355555216471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,10240,0.09623199701309204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,7168,0.030415998564826116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,5120,0.043621334764692515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,8192,0.07749155494901869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,7168,0.07766577932569715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,4096,0.035057776504092746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,6144,0.027266666293144226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,6144,0.06002044677734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,3584,0.03252355588807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,5120,0.024302222662501868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,3072,0.027676444914605882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,4096,0.020879111356205408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,5120,0.05083910955323113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,2560,0.02386666668785943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,4096,0.04218399855825636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,3072,0.017575111654069688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,3584,0.037463112009896174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,2048,0.018962666392326355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,3072,0.03361777795685662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,2560,0.015993777248594496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,1536,0.015067555838161044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,2560,0.02907555633121067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,2048,0.014190221826235453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,1024,0.011113777756690979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,2048,0.024491555160946313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,1536,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,16384,0.14856533209482828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,1024,0.01093777765830358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,1536,0.020071110791630216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,512,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,1024,0.01584533353646596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,768,0.010101333260536194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,256,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,768,0.013529777526855469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,512,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,128,0.004856889032655292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,512,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,256,0.009000889129108852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,64,0.004192000048028098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,32,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,128,0.009073778159088558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,256,0.01054666688044866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,128,0.010045333041085137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,64,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,4096,3584,0.01920000049802992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,4096,32,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,65536,0.4734791119893392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,51200,0.2957662211524116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,51200,0.18548888630337187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,16384,0.11880000432332356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,4096,768,0.008954666554927826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,65536,0.5690799819098579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,16384,0.057987557517157666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,51200,0.44449599583943683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,16384,0.1359857718149821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,10240,0.08177333407931857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,12288,0.0463120010164049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,12288,0.1028604441218906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,8192,0.06451644500096639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,10240,0.039855112632115684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,7168,0.05287999908129374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,10240,0.08636177910698785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,8192,0.03228088882234361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,6144,0.04571022258864509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,7168,0.029770665698581274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,8192,0.07014755407969157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,5120,0.037905779149797224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,7168,0.06293866369459364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,4096,0.03157066636615329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,6144,0.026778666509522334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,6144,0.05462577939033508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,5120,0.023631999890009563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,3584,0.025409777959187824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,4096,0.020239111449983384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,5120,0.04725155565473768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,65536,0.23979910214742026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,3584,0.018393778138690524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,4096,0.03886933459175958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,2560,0.020636444290479023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,2048,0.01620444489849938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,3072,0.01585422290696038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,3584,0.03477955526775784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,1536,0.01344622257683012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,12288,0.09127999676598443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,3072,0.03130133284462823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,1024,0.009726222190592024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,2560,0.026994667119450037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,2048,0.022474666436513264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,1536,0.011377777490350934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,768,0.007813333637184566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,1536,0.018360000517633226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,1024,0.010031999813185798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,512,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,1024,0.014793778459231058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,768,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,256,0.005058666484223472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,512,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,128,0.0047031111187405055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,256,0.008456000023418004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,256,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,128,0.009334222310119206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,128,0.008655110994974772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,64,0.004200889004601372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,3072,0.021783999270863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3584,32,0.004305777864323722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,64,0.009356444080670675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,2560,0.01442577772670322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,32,0.009316444396972656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3584,2048,0.01277955538696713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,65536,0.3270053333706326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,51200,0.25994933976067436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,65536,0.199891553984748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,51200,0.15925244490305582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,16384,0.0894355575243632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,768,0.01276177747382058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,12288,0.06392266352971394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3584,512,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,16384,0.051422221793068774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,65536,0.5163706673516167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,10240,0.06610666380988227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,16384,0.12096177207099067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,12288,0.039769778649012245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,8192,0.05303466651174757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,12288,0.09234133031633164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,7168,0.03860266672240363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,8192,0.028748446040683325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,10240,0.08003911044862536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,6144,0.03798133465978835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,7168,0.026763555076387193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,5120,0.032720890310075544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,7168,0.05729955434799194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,6144,0.02347466680738661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,4096,0.02426311042573717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,5120,0.020410666863123577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,5120,0.042509333954917065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,3584,0.02295911146534814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,4096,0.017664889494578045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,4096,0.03557866811752319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,3072,0.019371555911170114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,3584,0.01620444489849938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,3584,0.03211466802491082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,51200,0.4023555649651422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,2560,0.017086222767829895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,3072,0.015102222561836243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,2048,0.014523555835088095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,3072,0.028655999236636694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,1536,0.012080888781282636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,10240,0.03460089034504361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,2048,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,2560,0.024453333682484094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,1024,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,1536,0.01142133358452055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,8192,0.06437155273225573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,768,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,1024,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,1536,0.017616000440385606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,512,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,1024,0.013688888814714221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,256,0.004853333450025982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,768,0.01179377817445331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,512,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,128,0.004219555606444676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,512,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,256,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,64,0.003896000070704354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,256,0.009076444639099969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,128,0.008319999608728621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,3072,32,0.003903111235962974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,128,0.008767111433876885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,64,0.00868088917599784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,32,0.008656889200210571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,65536,0.3281493451860216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,2560,0.014101333088344999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,51200,0.22146666049957275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,65536,0.15443733003404406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,2048,0.020920000142521326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,16384,0.07562933365503947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,51200,0.12377156151665582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,12288,0.05430755681461758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,16384,0.04053599966896905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,3072,768,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,3072,6144,0.05018666717741224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,10240,0.05358844333224826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,51200,0.4386160108778212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,16384,0.10558666785558064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,8192,0.04368533359633552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,7168,0.03747377792994181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,12288,0.07909955581029256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,8192,0.022978666755888198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,10240,0.06734310918384127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,6144,0.033125334315829806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,7168,0.02159822152720557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,8192,0.06161066558625963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,5120,0.027386666999922857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,7168,0.04926310976346334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,6144,0.019034667147530448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,4096,0.02197155521975623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,6144,0.042908443344963916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,5120,0.016788444585270353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,3584,0.01979466610484653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,5120,0.03753777676158481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,4096,0.031239999665154353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,3072,0.01759111053413815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,3584,0.013575110998418597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,3584,0.02776266634464264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,2560,0.014797333214018079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,65536,0.4288604524400499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,3072,0.012727999852763282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,2048,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,3072,0.025045333637131587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,2560,0.01201866649919086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,1536,0.010339555641015371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,12288,0.03182844320933024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,2560,0.021293333835071985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,2048,0.010412444670995077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,1024,0.007869333028793335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,10240,0.028509332074059382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,768,0.0068604445291890045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,2048,0.018211555149820115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,1536,0.009531555904282464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,512,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,1024,0.00903555585278405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,1536,0.01518311103185018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,256,0.004254222330119875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,1024,0.012249777714411417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,512,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,768,0.010456889039940305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,128,0.003605333467324575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,512,0.009092444347010719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,256,0.007674666742483775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,64,0.0034328889515664843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2560,32,0.0035182221068276297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,4096,0.014872888724009195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,256,0.008636444807052612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,128,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,128,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,64,0.008341333104504479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2560,32,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,65536,0.2677706612481011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,51200,0.17789777119954428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,16384,0.0703164471520318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,12288,0.05269244313240051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,16384,0.03855289022127787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,65536,0.3765057722727458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,16384,0.09413422478569879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,51200,0.29734399583604604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,12288,0.030850665436850652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,8192,0.035785777701271906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,12288,0.06947466399934557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2560,768,0.008633777499198914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,7168,0.03074399961365594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,10240,0.060064891974131264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,8192,0.02234933276971181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,8192,0.049406223826938205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,7168,0.020485333270496793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,7168,0.04397955536842346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,5120,0.022608000371191237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,6144,0.018565333551830716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,6144,0.03837955660290188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,4096,0.018737778067588806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,65536,0.1452088885837131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,5120,0.016514667206340365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,3584,0.016400888562202454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,5120,0.03336000111367967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,4096,0.014550222290886773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,51200,0.11606310473548041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,3072,0.014110222458839417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,4096,0.027906666199366253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,2560,0.012327111429638334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,3584,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,3584,0.025037333369255066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,3072,0.012531555361217923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,2048,0.010392888552612728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,2560,0.011412444214026133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,3072,0.02233244478702545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,1536,0.008392000363932716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,10240,0.027871112028757732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,2560,0.01924977699915568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,1024,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,1536,0.009650666680600908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,768,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,1024,0.008749333520730337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,1536,0.014088888963063559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,512,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,1024,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,768,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,256,0.0038897775941424896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,768,0.009843555589516958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,512,0.008617777791288164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,128,0.003160888950030009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,256,0.007603555917739868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,256,0.007703111403518253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,64,0.0031884445084465873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,128,0.007541333635648091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,32,0.003206222214632564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,10240,0.04262755645645989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,64,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,32,0.007381333245171442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,65536,0.20586133003234863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,2048,0.010570666856235929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,2048,0.016491555505328707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,2048,6144,0.02625688910484314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,51200,0.18163022730085585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,65536,0.139573335647583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,16384,0.05968533621893989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,65536,0.3160711129506429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,51200,0.11244000328911676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,2048,512,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,12288,0.046819554434882275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,16384,0.03767377800411648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,10240,0.035451554589801364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,51200,0.2489688926272922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,16384,0.07929333051045735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,12288,0.029215110672844782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,8192,0.029623998536003962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,7168,0.02444533341460758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,10240,0.02660088903374142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,12288,0.06014133161968655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,8192,0.021163556310865615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,10240,0.05143644412358602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,6144,0.02236622240808275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,5120,0.018967999352349173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,7168,0.019557333654827543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,6144,0.017476444443066914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,7168,0.03743733300103082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,4096,0.015923556354310777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,6144,0.03337511089113023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,3584,0.014192889134089151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,5120,0.028528889020284016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,4096,0.013771555489963956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,3072,0.013085333009560903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,4096,0.023648000425762598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,3584,0.012798222402731577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,2560,0.011372444530328115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,3584,0.021524444222450256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,3072,0.012082666986518435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,2048,0.00906666699382994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,3072,0.019312888383865356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,2560,0.011063111325105032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,1536,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,2048,0.010067555639478896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,2048,0.014848889576064216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,1024,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,1536,0.009325332939624786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,1536,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,768,0.005522666705979242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,1024,0.008349333372380998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,1024,0.010062221851613786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,512,0.004591111093759537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,5120,0.01552177800072564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,768,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,256,0.0035342222286595237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,768,0.008627555436558193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,512,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,128,0.0031537777847713898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,512,0.007920888562997183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,256,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,64,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1536,32,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1536,128,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,256,0.007190222541491191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,128,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,64,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,65536,0.17479466067420113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,32,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,2560,0.016932444439993966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,65536,0.09389777978261311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,51200,0.13092354933420816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,16384,0.044693334235085376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,65536,0.22765334447224936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,12288,0.03415733244684007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,51200,0.07529599799050225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,10240,0.028375112348132666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,16384,0.02558311157756382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,51200,0.18058044380611846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,12288,0.021273778544531927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,16384,0.06006844176186455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,8192,0.021929777330822412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,10240,0.01924711134698656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,7168,0.019702222612169053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,8192,0.015558222929636637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,10240,0.03924177752600776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,6144,0.017818666166729398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,7168,0.01424088908566369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,8192,0.0326275560590956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,5120,0.015038222074508667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,7168,0.029389331738154095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,6144,0.01311822235584259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,4096,0.013120888835854001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,5120,0.011769777370823754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,6144,0.025898666845427618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,3584,0.011508444117175208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,5120,0.02236088944805993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,4096,0.010681777364677854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,3072,0.010484444598356882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,4096,0.01901511185699039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,3584,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,2560,0.009090666969617208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,3072,0.009759999811649323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,3584,0.01716800034046173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,2048,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,3072,0.01555999947918786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,2560,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,1536,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,2560,0.013728888498412238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,2048,0.008988444175985124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,2048,128,0.007329777710967594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,1024,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,2048,0.012094222009181976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,1536,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,768,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,1536,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,12288,0.0458586679564582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,1024,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,512,0.003919110943873723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,256,0.0031866667171319327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,768,0.00740799970097012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,1024,0.008650666309727563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,128,0.0028168888141711554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,512,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,768,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,64,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,256,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,1024,32,0.0028435554769304064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,1024,128,0.00674222202764617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,128,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,65536,0.12297333611382379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,64,0.005971555494599872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,32,0.006310222049554189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,51200,0.1060640017191569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,65536,0.09263910849889119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,16384,0.03415822320514255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,65536,0.22276711463928223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,51200,0.07447999715805054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1536,8192,0.04240622123082479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,12288,0.02436888880199856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,51200,0.1758568949169583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,10240,0.02128000060717265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,16384,0.025080889463424683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,16384,0.058111998769972056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,12288,0.02068800065252516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,8192,0.01741333305835724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,12288,0.04452088806364271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,7168,0.015980444020695157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,8192,0.015202666322390238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,512,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,8192,0.03166755702760484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,6144,0.01421511173248291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,7168,0.014160888062583076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,5120,0.012453333371215396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,7168,0.027985778119828966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,6144,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,4096,0.010037333601050908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,6144,0.024837333295080397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,5120,0.011781333221329583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,3584,0.009181333084901174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,4096,0.010693333215183683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,5120,0.021539555655585393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,3072,0.00831911133395301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,4096,0.018234666850831773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,3584,0.010255110760529837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,2560,0.007343110938866933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,3584,0.016536000702116225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,2048,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,3072,0.015167999598715039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,2560,0.013438222308953604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,1536,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,2048,0.008952889177534316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,2048,0.011611555185582904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,1024,0.004577777865860197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,1536,0.008273777862389883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,10240,0.03798044390148587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,768,0.004217777815130022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,1536,0.010051555931568146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,1024,256,0.006337777607970768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,1024,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,512,0.0035457776652442086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,1024,0.008286221987671321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,768,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,256,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,768,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,512,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,128,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,512,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,256,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,64,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,768,32,0.0028515555378463534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,256,0.006225777582989798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,128,0.006783999916579988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,128,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,65536,0.10364977518717448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,64,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,2560,0.009356444080670675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,51200,0.07988622453477648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,65536,0.09255644347932603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,65536,0.18352177408006456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,51200,0.07440000110202365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,10240,0.01904355486234029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,51200,0.14387288358476427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,16384,0.024238222175174292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,16384,0.048491554127799145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,10240,0.022481777601771887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,12288,0.01993955506218804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,12288,0.038035555018319026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,8192,0.018343999981880188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,10240,0.01834488908449809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,10240,0.03254844413863288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,7168,0.012688000169065265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,8192,0.014815110299322339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,8192,0.026986666851573523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,6144,0.01126933346192042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,7168,0.013799111048380533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,768,3072,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,5120,0.009768000079525841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,768,32,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,6144,0.012637333737479316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,16384,0.02924799919128418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,4096,0.008083555433485243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,6144,0.02166844407717387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,5120,0.011363555159833698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,3584,0.007833778030342525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,5120,0.01881511178281572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,4096,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,12288,0.026484444737434387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,3584,0.010024888647927178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,4096,0.01592177814907498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,2560,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,3584,0.014769777655601501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,3072,0.00963466697269016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,2048,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,3072,0.013461333182122974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,1536,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,2560,0.012072000238630505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,1024,0.003559999995761447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,2048,0.010392888552612728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,1536,0.008010666403505538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,768,0.003181333343187968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,1536,0.009148444566461775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,1024,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,512,0.003175111073586676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,1024,0.007672888537247975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,768,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,7168,0.024604444702466328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,256,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,512,0.006908444480763541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,128,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,512,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,256,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,64,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,256,0.005728888842794631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,128,0.006607999818192587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,3072,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,512,32,0.0022293333378103045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,128,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,65536,0.050456000698937305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,64,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,2560,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,65536,0.09200000100665623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,512,2048,0.008612444831265343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,65536,0.14085866345299616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,51200,0.07419821951124403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,51200,0.11115377479129368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,16384,0.02347911066479153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,12288,0.014208000567224292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,10240,0.013069333301650153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,16384,0.03805689016977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,12288,0.0199937770764033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,8192,0.01315822203954061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,768,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,12288,0.029887109994888306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,10240,0.018276444739765592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,7168,0.011173333558771344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,10240,0.025754666990704004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,6144,0.00979466653532452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,8192,0.014557333456145393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,7168,0.013623110949993134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,8192,0.021701332595613267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,5120,0.00868977771864997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,7168,0.01960355540116628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,6144,0.012554667062229581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,6144,0.017620444297790527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,5120,0.011433777709801992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,51200,0.04064355625046624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,3584,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,5120,0.015625778171751235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,512,32,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,4096,0.010325333310498131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,4096,0.01348533315791024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,3584,0.010145777629481422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,2560,0.005375111268626318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,3584,0.012440000143316058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,2048,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,2560,0.008983111215962304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,3072,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,1536,0.004013333469629288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,2560,0.01038577738735411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,2048,0.008362666600280339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,1024,0.0032017777363459268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,2048,0.009101333717505137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,1536,0.008013333711359236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,1536,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,768,0.0031475555151700974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,1024,0.007663110891977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,1024,0.0069919998447100324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,512,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,768,0.00699733363257514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,4096,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,256,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,768,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,512,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,128,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,512,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,256,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,64,0.0021582223061058256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,256,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,16384,0.01685066686736213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,128,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,32,0.002114666625857353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,256,3072,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,128,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,51200,0.03386044502258301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,64,0.005212444398138259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,256,32,0.005183111048407025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,65536,0.09161600139405991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,16384,0.010384000009960597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,65536,0.14026578267415366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,51200,0.07368622223536174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,12288,0.009182222187519073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,51200,0.11053599913914998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,10240,0.010942222343550788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,12288,0.019665777683258057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,16384,0.03759555684195624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,8192,0.00959022260374493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,12288,0.02958844436539544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,10240,0.018239999810854595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,7168,0.009033777647548253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,10240,0.0251217782497406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,8192,0.014681778020328946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,6144,0.008921778036488427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,256,3072,0.006088888893524806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,7168,0.013467555244763693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,5120,0.008656889200210571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,7168,0.019359999232821994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,6144,0.012449777788586087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,65536,0.04203377829657661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,4096,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,6144,0.01753777762254079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,5120,0.011372444530328115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,3584,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,4096,0.010390222072601318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,3072,0.005680888891220093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,3584,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,4096,0.013355555633703867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,3584,0.012392000191741519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,2560,0.005197333378924264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,2048,0.004537777768241035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,3072,0.011423110961914062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,2560,0.010057777994208867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,2048,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,1536,0.003847111016511917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,2048,0.00908799966176351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,1024,0.0031715554909573663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,1536,0.00794222205877304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,8192,0.021293333835071985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,768,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,1536,0.008203555312421586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,1024,0.00693866651919153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,512,0.002579555536309878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,768,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,768,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,256,0.0024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,512,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,5120,0.015480889214409722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,512,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,128,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,256,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,64,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,128,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,256,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,3072,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,128,32,0.0022400000856982335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,128,0.005105777747101254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,65536,0.03716266817516751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,2560,0.008967111508051554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,51200,0.02955289019478692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,16384,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,12288,0.00978133330742518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,32,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,10240,0.008799999952316284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,65536,0.14002133740319148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,8192,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,51200,0.11021066374248928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,7168,0.008683555656009251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,12288,0.028992888000276353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,6144,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,10240,0.02507200009293026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,5120,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,8192,0.021203555994563635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,4096,0.006721777634488211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,6144,0.017440888616773818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,5120,0.015456888410780164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,3584,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,16384,0.023175110419591267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,4096,0.013239111337396832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,3072,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,2560,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,3584,0.012396444049146442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,2048,0.004513777792453766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,128,64,0.005151999907361137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,1536,0.0038853333228164246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,3072,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,1024,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,2560,0.01016266644001007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,16384,0.0373564428753323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,768,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,512,0.0025271111064487034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,2048,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,256,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,1536,0.008213332957691616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1537,128,1024,0.007442666424645319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,128,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,1024,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,64,0.0019048889064126545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,64,32,0.0018800000349680583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,768,0.0064275554484791225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,512,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,65536,0.0410773323641883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,256,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,51200,0.030530667967266504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,128,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,16384,0.009323555562231276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,12288,0.011642666326628791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,32,0.004951111144489712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,10240,0.01053244454993142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,65536,0.13965600066714817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,16384,0.03714222378200955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,8192,0.009338666995366415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,7168,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,10240,0.02514666650030348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,6144,0.012422222230169507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,8192,0.021235555410385132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,5120,0.011041777829329172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,6144,0.01756355497572157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,4096,0.007851555943489075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,5120,0.015338665909237333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,3584,0.007252444823582967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,7168,0.01937866707642873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,3072,0.006567111031876669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,2560,0.00573600000805325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,4096,0.013439999686347114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,64,64,0.004981333182917701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,2048,0.004960000101062986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,1536,0.0037360000941488477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,3584,0.012271111210187277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,3072,0.01145066652033064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,1024,0.0031555555760860443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,2560,0.010320000350475311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,768,0.0028515555378463534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,512,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,51200,0.11020888884862264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,2048,0.008974221845467886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,256,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,1536,0.008289777570300633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,12288,0.02900000082121955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,128,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,64,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,1024,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1537,32,32,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,768,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,512,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,256,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,128,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,7168,0.01941688855489095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,64,0.005049777941571342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1537,32,32,0.004973333328962326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,16384,0.5850862397087945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,16384,1.9252230326334636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,16384,2.127471923828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,12288,1.5355004204644098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,51200,2.125282711452908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,16384,1.881259494357639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,12288,0.44789690441555446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,12288,1.5384239620632594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,10240,0.3933440049489339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,10240,1.3218373192681205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,12288,1.3486400180392797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,10240,1.2453911039564345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,8192,1.052276399400499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,8192,0.32553333706325954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,51200,6.2689420912000875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,10240,1.101178699069553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,8192,1.104323599073622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,7168,0.8641022046407064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,7168,0.2733244366115994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,8192,0.8747875425550672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,7168,0.8507253328959147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,6144,0.7282693121168348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,6144,0.25358398755391437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,5120,0.533014244503445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,6144,0.7483875486585828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,51200,7.986676534016927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,5120,0.755432022942437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,7168,0.739833778805203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,6144,0.6233555475870768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,4096,0.41352089246114093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,4096,0.17267466915978325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,3584,0.4226693312327067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,4096,0.49441157446967227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,5120,0.5419795778062608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,4096,0.4365528954399957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,3584,0.15996532970004612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,3584,0.4294275442759196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,3072,0.35289422671000165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,5120,0.2125422159830729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,2560,0.2875235610538059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,3072,0.13746578163570827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,3072,0.3670586744944255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,3072,0.34155644310845273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,2560,0.1210986640718248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,2560,0.3086080021328396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,2048,0.24768087599012586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,2048,0.1036248869366116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,1536,0.17596088515387642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,2560,0.28856976826985675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,2048,0.25295999315049916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,2048,0.23342755105760363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,1024,0.11461244689093696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,1536,0.08408888843324448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,1536,0.20039733250935873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,768,0.10105955600738525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,1024,0.06623822450637817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,1024,0.15230845080481634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,512,0.0674826635254754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,1024,0.14401511351267496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,3584,0.3845991028679742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,768,0.058576001061333545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,768,0.12204355663723415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,51200,6.036888122558594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,768,0.11825156211853027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,256,0.053949332899517484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,512,0.05322755707634819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,128,0.04258844587537977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,512,0.10509422090318467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,256,0.09408889214197795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,512,0.0985777775446574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,256,0.04891733328501383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,64,0.04003289010789659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,65536,128,0.04670400089687771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,256,0.0980151096979777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,128,0.08841777510113186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,128,0.05864533450868395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,65536,32,0.039008888933393694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,64,0.08782044384214614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,65536,32,0.08888000249862671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,65536,1536,0.20390933089786104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,51200,1.742410659790039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,65536,2.20851198832194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,51200,4.914833068847656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,16384,1.8890914916992188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,16384,1.8231822119818792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,16384,0.5106666882832845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,51200,5.941890716552734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,65536,6.672968970404731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,12288,1.3918150795830622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,12288,1.2512426376342773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,12288,0.3733697732289632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,10240,0.9053857591417102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,16384,1.3860017988416884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,12288,1.0496124691433377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,65536,8.515402052137587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,10240,0.9791128370496961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,10240,0.3088151084052192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,8192,0.7704408963521322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,51200,4.647221459282769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,8192,0.25289866659376353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,8192,0.7899040116204156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,10240,0.8708097669813367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,7168,0.6825511190626355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,7168,0.22382044792175293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,6144,0.4851973321702745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,7168,0.6757377518547906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,6144,0.19395200411478677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,8192,0.6711786588033041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,6144,0.6253457599216038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,7168,0.5796960194905599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,5120,0.16779288980695936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,4096,0.38189421759711367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,5120,0.4747386508517795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,6144,0.5106639862060547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,4096,0.14870399898952907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,4096,0.4091991053687201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,5120,0.4247475465138753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,65536,6.051594628228082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,3584,0.3357715606689453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,3584,0.12363021903567845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,4096,0.3498426543341742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,3072,0.2994942135281033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,2560,0.25154222382439506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,3072,0.11058311329947577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,3584,0.3122284412384033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,3072,0.32774755689832896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,3072,0.27070310380723744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,2560,0.2819937864939372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,2048,0.19006133079528809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,2560,0.0996862252553304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,2560,0.2661831114027235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,3584,0.2924559911092122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,1536,0.15994755427042642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,2048,0.08381688594818115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,2048,0.19957866933610705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,2048,0.1892951064639621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,1024,0.10125066836675008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,1536,0.06726044416427612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,768,0.09157333109113906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,1024,0.05335288908746508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,1024,0.12047021918826634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,1536,0.14944444762335882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,512,0.0603075557284885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,1024,0.11520000298817952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,768,0.047672000196244985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,768,0.09705599811342026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,256,0.03916710946294997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,512,0.04299911194377475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,512,0.0835262205865648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,512,0.07972177531984118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,128,0.03353777858946059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,256,0.07479555739296807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,256,0.03975288735495673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,256,0.06242044766743978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,64,0.030689779255125258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,51200,128,0.03819377885924445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,128,0.07066488928265043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,128,0.047891553905275136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,1536,0.16815022627512613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,64,0.07019378079308404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,51200,32,0.07072799735599093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,5120,0.449727111392551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,51200,768,0.09768799940745036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,51200,32,0.029267556137508813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,65536,0.6884933577643501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,51200,1.4844364590115016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,65536,2.016487969292535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,16384,0.5946782430013021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,51200,0.5518310864766439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,16384,0.18203288979000518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,51200,1.9205716451009114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,12288,0.40921688079833984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,16384,0.5391777886284722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,65536,2.637080086602105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,10240,0.27723643514845103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,12288,0.13392799430423313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,12288,0.3963350984785292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,16384,0.45401864581637913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,10240,0.3297235435909695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,10240,0.10754755470487808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,8192,0.27201599544949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,51200,1.4735928641425238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,8192,0.0880533324347602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,10240,0.27731556362575954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,7168,0.24544533093770346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,65536,2.044945822821723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,6144,0.20708355638715956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,7168,0.2435164451599121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,8192,0.2305920124053955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,7168,0.07682044638527764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,7168,0.20229600535498726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,5120,0.1710595554775662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,6144,0.0701999995443556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,6144,0.21912177403767905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,4096,0.12569955984751383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,5120,0.05965422259436714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,6144,0.19497599866655138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,12288,0.33814311027526855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,3584,0.11439732710520427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,4096,0.050208889775806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,5120,0.14888532956441244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,4096,0.1308737728330824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,3072,0.08876711130142212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,8192,0.26538578669230145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,3584,0.04506488972240024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,2560,0.08603199985292222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,3584,0.11144444677564834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,3072,0.04015377826160855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,3072,0.09956266482671101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,3584,0.11320443948109944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,2048,0.06452177630530463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,3072,0.09761510954962836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,2560,0.03548177745607164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,2560,0.08499111069573297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,1536,0.04921066761016846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,2560,0.08584800031450059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,2048,0.030417776770061914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,1024,0.03421866562631395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,2048,0.07208977805243598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,1536,0.025069332785076563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,1536,0.05570933553907606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,1536,0.05975288814968533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,768,0.025247111916542053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,5120,0.15845510694715711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,1024,0.020689777202076383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,1024,0.04258222381273905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,512,0.018722666634453666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,1024,0.04710133208168877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,768,0.01864355636967553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,4096,0.1250844399134318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,768,0.03475111060672336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,256,0.01457333399189843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,768,0.04035822219318814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,512,0.017108443710539076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,512,0.02846222122510274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,128,0.014160000615649752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,256,0.015485333071814643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,512,0.0343191126982371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,256,0.0255386663807763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,64,0.011952000359694162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,16384,32,0.012362666428089142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,16384,128,0.015489778584904142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,128,0.024689777029885188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,128,0.020977778567208186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,2048,0.07083377573225233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,64,0.0244195560614268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,16384,32,0.024907555845048692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,65536,0.5044906404283311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,16384,256,0.027315555347336665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,51200,1.0769839816623263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,65536,1.5576826731363933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,16384,0.3621999952528212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,51200,1.448097758822971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,51200,0.40144801139831543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,65536,1.7737307018703883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,16384,0.12525510787963867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,12288,0.27472178141276044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,16384,0.4414382245805528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,10240,0.23451378610399035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,12288,0.3042337894439697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,16384,0.34412890010409886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,12288,0.25984621047973633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,10240,0.2413395510779487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,8192,0.19880710707770455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,10240,0.08187377452850342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,65536,1.4947084850735133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,10240,0.21933689382341173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,8192,0.19086221853892008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,8192,0.06985421975453694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,8192,0.1758915583292643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,7168,0.06019644604788887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,51200,1.1444951163397896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,7168,0.19806399610307482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,5120,0.1191600031322903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,7168,0.15936177306705052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,6144,0.06115111377504137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,6144,0.14587644735972086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,12288,0.09688266780641343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,6144,0.13468978140089247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,5120,0.04850133260091146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,5120,0.12504177623324925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,5120,0.11771466996934678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,3584,0.0840853320227729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,4096,0.03957510987917582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,4096,0.09899644719229804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,3072,0.07525510920418634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,4096,0.09648533662160237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,7168,0.17060711648729113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,3584,0.08783555693096584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,3584,0.08881511290868123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,2560,0.05758577585220337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,6144,0.14770844247606066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,3072,0.030392888519499037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,3072,0.07659377654393514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,2048,0.04874933428234524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,2560,0.02749155627356635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,3072,0.07723022169537015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,2560,0.06691644589106242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,1536,0.03699288765589396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,2560,0.06853244702021281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,2048,0.05494844582345751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,1024,0.023920888702074688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,2048,0.05774933099746704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,4096,0.09753244453006321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,1536,0.02036977807680766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,768,0.018911111685964797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,1536,0.04453511039415995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,1536,0.04753422074847751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,1024,0.016952888833151925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,1024,0.03386755453215705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,512,0.01460444430510203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,768,0.015196444259749519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,768,0.027828445037206013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,3584,0.035445332527160645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,256,0.01087288889620039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,768,0.03303111261791653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,512,0.014128888646761576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,512,0.02277333372169071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,128,0.010249777800507015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,512,0.02819288770357768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,256,0.013190222283204397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,256,0.020071110791630216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,128,0.01311377767059538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,256,0.02162933349609375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,128,0.019595555133289762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,32,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,128,0.0173662222094006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,64,0.01965688831276364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,12288,2048,0.023756444454193115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,12288,32,0.01958044370015462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,12288,1024,0.03858133488231235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,65536,0.46869775984022355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,12288,64,0.008951110972298516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,65536,1.3118444018893771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,51200,0.9818106757269965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,16384,0.3337582217322455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,65536,1.4453484217325847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,51200,0.3545937803056505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,12288,0.22939554850260416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,16384,0.10644000106387669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,51200,1.1402097278171115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,12288,0.2691306538052029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,10240,0.19022844897376168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,16384,0.3066551155514187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,12288,0.23024090131123862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,65536,1.292616844177246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,10240,0.0726977785428365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,10240,0.20208888583713105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,8192,0.1669973267449273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,10240,0.18994222746955025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,7168,0.1369653277926975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,51200,0.9888088438245984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,8192,0.1912053293652005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,8192,0.15604178110758463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,7168,0.06016888883378771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,16384,0.37374401092529297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,7168,0.14292621612548828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,5120,0.09579466448889838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,6144,0.04634399877654182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,7168,0.14827466011047363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,12288,0.08185688654581706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,6144,0.12362489435407852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,4096,0.08623111248016357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,6144,0.12432977888319229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,5120,0.03938222262594435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,5120,0.10381688674290974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,3584,0.07816621992323133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,5120,0.10359377993477716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,4096,0.08448977602852716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,3072,0.06787821981641981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,4096,0.08712889088524713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,3584,0.030024889442655776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,3584,0.07468177874883015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,8192,0.0583768884340922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,3584,0.07768088579177856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,3072,0.027154665854242113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,3072,0.0651457773314582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,6144,0.12916710641649035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,2048,0.04305600126584371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,2560,0.05707199705971611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,2560,0.06036000119315254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,2048,0.021203555994563635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,1536,0.03455288873778449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,2048,0.04732266730732388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,2048,0.05194133188989428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,1024,0.024506666594081458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,1536,0.01761511133776771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,1536,0.03843911157713996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,1536,0.04261777798334757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,768,0.020434666011068556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,1024,0.02920444475279914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,1024,0.03497955534193251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,2560,0.053399999936421715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,512,0.01776444415251414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,768,0.013948443863126965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,768,0.029870222012201946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,2560,0.024664888779322307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,256,0.01350400017367469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,512,0.019862223002645705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,512,0.02511288887924618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,256,0.017549332645204332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,128,0.00792266676823298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,256,0.01182755579551061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,256,0.020890666378868952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,4096,0.034295111894607544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,128,0.011952000359694162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,128,0.01721600029203627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,32,0.010728889041476779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,128,0.016693333784739178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,64,0.01719911065366533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,768,0.02420444455411699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,10240,32,0.01716088917520311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,10240,3072,0.07002311282687716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,512,0.01293244461218516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,65536,0.3746524386935764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,65536,0.9478577507866753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,10240,64,0.010688888529936472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,51200,0.7992186546325684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,10240,1024,0.01516355574131012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,16384,0.28323109944661456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,51200,0.2880604532029894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,65536,1.1806364059448242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,51200,0.9203306833902994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,12288,0.20372177494896782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,16384,0.0929715567164951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,16384,0.28071022033691406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,16384,0.25637955135769314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,10240,0.16480710771348742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,12288,0.07333688603507148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,12288,0.1983040041393704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,65536,1.0773404439290364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,12288,0.20563199785020617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,10240,0.0613973339398702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,10240,0.16297422515021429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,7168,0.11687200599246556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,51200,0.8136222097608777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,10240,0.16715733210245767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,8192,0.14141333103179932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,8192,0.13672355810801187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,7168,0.043805334303114146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,6144,0.09533955653508504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,7168,0.1167351139916314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,5120,0.08253777689403959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,7168,0.12828177875942653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,6144,0.039452443520228066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,6144,0.1020853320757548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,6144,0.10619288682937622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,5120,0.033752888441085815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,5120,0.08614222208658855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,5120,0.09232711129718357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,8192,0.11377599504258896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,3584,0.05885599719153511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,4096,0.02882666720284356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,4096,0.07013333506054349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,3072,0.05044710967275831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,4096,0.07645777861277263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,3584,0.02630666726165348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,8192,0.04987644486957126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,3584,0.0620595547888014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,2560,0.04366311099794176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,3584,0.06759822368621826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,3072,0.023804444405767653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,3072,0.054690665668911405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,2048,0.03666844301753574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,3072,0.06090933084487915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,2560,0.021719111336602107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,1536,0.02572711143228743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,2560,0.053426666392220386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,2048,0.018960888187090557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,2048,0.03970400161213345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,1024,0.01867022282547421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,2048,0.045951998896069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,4096,0.0666746695836385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,768,0.015484443969196744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,1536,0.03200799889034695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,1536,0.03842577669355605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,1024,0.024880000286632117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,512,0.011763555308183035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,1024,0.030624889665179785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,768,0.012552888856993781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,768,0.020378667447302077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,256,0.011389333340856763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,768,0.02681866619322035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,512,0.011776888536082374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,512,0.01703733371363746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,128,0.009489778015348647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,512,0.022154667311244543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,256,0.011013333168294696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,256,0.015164444843928019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,64,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,128,0.01109155515829722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,128,0.01481599940194024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,2560,0.047262221574783325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,128,0.014988443917698331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,64,0.014815110299322339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,1536,0.01628177695804172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,8192,32,0.014519110321998596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,8192,1024,0.013803555733627744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,65536,0.3402142259809706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,65536,0.8653590944078234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,51200,0.714445326063368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,8192,256,0.01897333396805657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,8192,32,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,65536,1.010873794555664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,16384,0.2671857674916585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,51200,0.8088169097900391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,51200,0.25653065575493705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,16384,0.08052000072267321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,16384,0.23491732279459634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,12288,0.2001955509185791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,10240,0.15970667203267416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,12288,0.06281689140531752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,12288,0.18300444549984404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,65536,0.9459270901150174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,12288,0.19180088573031953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,8192,0.1283822192086114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,10240,0.15188799964057073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,10240,0.05393866697947184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,7168,0.10925777753194173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,8192,0.043043557140562266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,10240,0.14509688483344182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,8192,0.1350542174445258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,51200,0.7395671208699545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,6144,0.09488799836900498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,7168,0.10445511341094971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,7168,0.0385155545340644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,5120,0.08232888910505506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,7168,0.10784177647696601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,6144,0.034056001239352755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,6144,0.09370933638678657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,6144,0.09980622265073989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,4096,0.0625422199567159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,5120,0.029881778690550063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,5120,0.07681333356433444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,3584,0.05882755253050062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,4096,0.025576000412305195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,5120,0.08161155382792155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,4096,0.06509333186679415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,16384,0.22368534406026205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,4096,0.06786044438680013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,3072,0.04876800047026741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,3584,0.02349066734313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,2560,0.041414221127827965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,3072,0.021500445074505273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,3072,0.049599998527103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,3072,0.05489155650138855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,2048,0.033688889609442815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,2560,0.019564444820086162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,8192,0.12710489167107478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,2560,0.0480275551478068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,1536,0.026541334059503343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,2048,0.017196445001496207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,2048,0.035982221364974976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,2048,0.04112177756097581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,1536,0.029603554142846003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,1536,0.014648889501889547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,1536,0.03408266769515143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,1024,0.022568888134426538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,1024,0.012597333225939008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,768,0.015867556134859722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,1024,0.02749066717094845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,3584,0.05664889017740885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,512,0.01108000013563368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,3584,0.06092888779110379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,768,0.011755555868148804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,768,0.019007111589113872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,256,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,768,0.023548444112141926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,512,0.011056889262464313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,512,0.015446222490734525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,128,0.007095110913117726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,512,0.020220445262061227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,256,0.010241777532630498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,256,0.014114666316244336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,64,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,256,0.017064000169436138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,7168,128,0.01035644445154402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,32,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,128,0.013638222383128272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,7168,1024,0.01902222302224901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,7168,128,0.014144000079896716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,64,0.013502221968438892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,32,0.013602666556835175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,51200,0.4605146514044867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,65536,0.2699475553300646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,65536,0.7103146447075738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,7168,2560,0.04339111182424757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,16384,0.18428444862365723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,65536,0.9049395455254449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,51200,0.7049999766879611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,16384,0.0736480024125841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,16384,0.272360881169637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,12288,0.12099111080169678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,10240,0.10542577505111694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,16384,0.18927555614047578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,51200,0.5947137938605415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,65536,0.7507902251349555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,12288,0.1532035536236233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,8192,0.09606489208009507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,12288,0.14977867073482937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,10240,0.12992089324527317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,10240,0.12549155288272434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,7168,0.0846595565478007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,8192,0.11390754911634658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,7168,0.03394044438997904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,51200,0.21477599938710532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,7168,0.09170933564503987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,6144,0.07630399862925212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,6144,0.029855112234751385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,5120,0.06367288695441352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,12288,0.05501066644986471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,6144,0.079276442527771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,6144,0.08029777473873563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,4096,0.045520888434516064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,10240,0.04964888758129544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,5120,0.025849777791235182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,5120,0.06842133071687487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,5120,0.0693804423014323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,3584,0.04487999942567614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,4096,0.02253599961598714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,8192,0.03674133287535773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,8192,0.10393155283398098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,3072,0.036672890186309814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,4096,0.05532177620463901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,4096,0.05840533309512668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,3584,0.020610666937298246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,2560,0.03238133258289761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,3584,0.0495013329717848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,7168,0.09162844551934136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,3584,0.05351377858055962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,2048,0.024550222688251074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,3072,0.04839200112554762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,2560,0.01742755538887448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,2560,0.03796622157096863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,1536,0.018567111757066514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,2560,0.041647109720442034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,2048,0.0151591118839052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,2048,0.032419555717044406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,1024,0.014129777749379476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,2048,0.03630933496687148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,1536,0.013450666434235044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,1536,0.026365333133273657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,768,0.011383111278216044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,1536,0.030438221163219873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,1024,0.01145244472556644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,1024,0.020132443971104093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,512,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,1024,0.024326221810446844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,768,0.010738666686746808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,768,0.01702666613790724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,256,0.006497777584526274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,768,0.021623111433453027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,512,0.014142221874660917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,512,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,128,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,3072,0.018965333700180054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,256,0.009376889301670922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,256,0.012729778057999082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,64,0.005486222190989389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,256,0.016505777835845947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,6144,32,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,128,0.012423111332787408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,128,0.013262222210566202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,64,0.012376888758606382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,32,0.012277333272827996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,65536,0.6394782066345215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,65536,0.2561448944939507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,51200,0.42297689119974774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,65536,0.7610658009847006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,6144,512,0.01905866629547543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,16384,0.15792266527811685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,51200,0.201928006278144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,6144,128,0.009536888864305284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,16384,0.06387288702858819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,12288,0.12470666567484538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,16384,0.17282844914330375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,65536,0.7446257803175184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,51200,0.6597929000854492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,10240,0.11256533198886448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,12288,0.05030044582155016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,51200,0.5745564566718208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,12288,0.1302071147494846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,8192,0.09142044517729019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,10240,0.12010222011142307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,10240,0.1314800050523546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,8192,0.03542133172353109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,8192,0.08968266513612534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,7168,0.07655644416809082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,6144,3072,0.04396977689531115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,8192,0.10537066724565293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,6144,0.05966666671964857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,7168,0.03277422322167291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,7168,0.07938488986757067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,7168,0.08882400062349106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,5120,0.05750311083263821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,6144,0.028757333755493164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,6144,0.06895111004511516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,6144,0.07826755444208781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,4096,0.04524533285035027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,5120,0.025243555506070454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,16384,0.19972533649868437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,4096,0.02169955604606205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,3584,0.03947733177079095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,4096,0.04778133498297798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,12288,0.1430551078584459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,4096,0.05729866690105862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,3072,0.03424088822470771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,3584,0.019934222102165222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,3584,0.042950222889582314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,3584,0.051706665092044406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,3072,0.018202667435010273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,3072,0.037800888220469155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,3072,0.04633333285649618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,2048,0.0244862222009235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,2560,0.016521778371598985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,2560,0.04051466782887777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,2048,0.028322666883468628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,1536,0.019175110591782462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,2048,0.014511111709806653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,2048,0.03501066565513611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,5120,0.05877600113550822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,5120,0.06793777810202704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,1536,0.013004444539546967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,1536,0.022924444741672937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,10240,0.04359466830889384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,1536,0.028743111424975928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,1024,0.01756000022093455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,768,0.012039110892348819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,1024,0.011066666907734342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,2560,0.030227555168999568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,1024,0.02312266661061181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,512,0.009117333425415887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,768,0.010399999717871347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,768,0.015033778217103748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,256,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,512,0.009757333331637913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,768,0.020632000433074105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,512,0.012423999607563019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,128,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,512,0.01789422167672051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,256,0.009367110828558603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,256,0.011508444117175208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,64,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,256,0.015692444311247934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,128,0.011047999891969891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,32,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,5120,128,0.012111999922328524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,5120,1024,0.01463733282354143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,64,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,32,0.011031111081441244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,65536,0.4340248902638753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,65536,0.2071839968363444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,51200,0.35152623388502335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,5120,2560,0.033062222931120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,65536,0.6131262249416775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,16384,0.1281599998474121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,51200,0.16377422544691297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,51200,0.5019760131835938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,65536,0.5869786474439833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,16384,0.14341155687967935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,16384,0.05222400029500326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,12288,0.09183644586139256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,10240,0.08610666460461086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,16384,0.1532177792655097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,12288,0.04177777634726631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,51200,0.4668524530198839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,12288,0.11530133088429768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,8192,0.0786000026596917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,10240,0.0362115568584866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,10240,0.09382577737172444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,7168,0.068159110016293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,8192,0.030898667044109766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,10240,0.10227021906110977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,8192,0.07563110854890612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,8192,0.08367644415961371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,6144,0.05437155564626058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,7168,0.02793866727087233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,5120,0.04777955677774218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,7168,0.06752977768580119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,6144,0.02383022175894843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,7168,0.07365510861078899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,6144,0.05864355299207899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,4096,0.03728355632887946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,5120,0.02143999934196472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,6144,0.06634400288263957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,5120,128,0.009328888522254096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,5120,0.05102133419778612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,3584,0.03153333399030898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,5120,0.05704622136221992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,4096,0.019034667147530448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,3072,0.027455111344655354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,4096,0.04748177859518263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,3584,0.03757066528002421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,3584,0.04356000158521864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,3072,0.015580443872345818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,3072,0.03347377644644843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,3072,0.039284444517559476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,2048,0.020282665888468426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,2560,0.014454222387737699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,2560,0.028984887732399836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,2560,0.034501334031422935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,1536,0.013952889376216464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,2048,0.013628444737858243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,2048,0.02455377744303809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,2048,0.03039466672473484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,1536,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,1536,0.02017600006527371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,1536,0.024620445238219366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,4096,0.0418177776866489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,1024,0.01551999979548984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,768,0.008691555923885768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,3584,0.01687466601530711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,1024,0.020807110601001315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,512,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,2560,0.02360800074206458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,768,0.00941244430012173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,12288,0.11769510640038384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,256,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,768,0.018266666266653273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,512,0.00886044485701455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,512,0.011409777734014722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,128,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,512,0.01754044493039449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,256,0.008619555168681675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,256,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,256,0.014710222681363424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,64,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,128,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,128,0.008414222134484185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,32,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,4096,128,0.011968000067604912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,4096,1024,0.010797333386209277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,64,0.009768889182143742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,32,0.009647111097971598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,4096,1024,0.010077333284748925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,65536,0.4335297743479411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,65536,0.19547200202941895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,51200,0.33966933356391055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,65536,0.5621235635545518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,16384,0.11384088463253444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,51200,0.15383466084798178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,51200,0.4417751100328233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,65536,0.5724150869581435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,12288,0.0941244430012173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,16384,0.05113510953055488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,16384,0.13043910927242702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,51200,0.4568257861667209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,16384,0.15183822313944498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,10240,0.07945599820878771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,12288,0.039981332090165876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,12288,0.1003555523024665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,8192,0.06848355796602038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,10240,0.03499822152985467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,12288,0.11687555578019883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,10240,0.08556266625722249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,8192,0.028241776757770117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,8192,0.06885333193673028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,10240,0.09964533646901448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,8192,0.08252977662616305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,6144,0.04807644420199924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,7168,0.025855110751258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,4096,768,0.013286222186353473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,7168,0.06070399946636624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,5120,0.04270844327078926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,6144,0.023223999473783705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,7168,0.07372977998521593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,6144,0.05331644415855408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,4096,0.03416444526778327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,5120,0.020633776982625324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,3584,0.03065511253145006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,4096,0.017960889471901786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,5120,0.056332442495557994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,4096,0.0384897788365682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,4096,0.047200000948376127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,3072,0.025948445002237957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,3584,0.01664888858795166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,3584,0.034633778863483004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,2560,0.022795556320084467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,3584,0.042697777350743614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,3072,0.015311110350820752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,3072,0.030830221043692693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,2048,0.0188755558596717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,7168,0.05717866950564914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,3072,0.03828800055715773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,2560,0.02664711078008016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,1536,0.0151582227812873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,2560,0.03397511111365424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,2048,0.012416889270146688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,2048,0.02273688879277971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,1024,0.011200000014570026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,2048,0.0289066665702396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,1536,0.018593778212865193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,1536,0.025066667132907446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,5120,0.0457839998934004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,768,0.007879999776681265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,1024,0.014426666829321118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,1024,0.020082667469978333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,512,0.0064328888224230866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,768,0.009309333231714036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,768,0.012493333054913415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,768,0.017870222528775532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,256,0.005192000004980299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,512,0.008834666675991481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,512,0.01070577816830741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,512,0.015814221567577787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,128,0.004518222063779831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,2560,0.013853333062595792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,256,0.00831200016869439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,256,0.009687110781669617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,64,0.0041582224269707995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,256,0.014481777946154276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,128,0.00833777752187517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3584,32,0.004469333423508538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,128,0.009404444032245213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,128,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,1536,0.011063999599880643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,64,0.009299555586444007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3584,32,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3584,6144,0.06526488727993436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,65536,0.32822932137383354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,65536,0.16036532984839544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,51200,0.25808265474107533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,16384,0.10406400097741021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,65536,0.4795066515604655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,51200,0.12610222233666316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,51200,0.3853448761834039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,16384,0.11617510848575169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,16384,0.040935112370385066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,51200,0.3455137676662869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,12288,0.0720577769809299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,16384,0.11887555652194554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,12288,0.03158400124973721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,12288,0.08762133121490479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,12288,0.09337600072224934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,10240,0.0737537807888455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,8192,0.0484844446182251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3584,1024,0.009988444546858469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,10240,0.08044177956051297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,7168,0.041492445601357356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,8192,0.023016888234350417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,8192,0.060250666406419545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,8192,0.06684799989064534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,6144,0.03590933481852213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,7168,0.02094399929046631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,7168,0.05414399835798475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,7168,0.06000622113545736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,65536,0.44741423924763996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,6144,0.019125334090656705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,5120,0.029599110285441082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,6144,0.047520889176262744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,4096,0.024672889047198828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,6144,0.05207911133766174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,5120,0.01715555621518029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,5120,0.04056088791953193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,3584,0.02208711041344537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,4096,0.014949333336618213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,4096,0.03377866744995117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,10240,0.027942223681343928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,3072,0.019547555181715224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,4096,0.038387556870778404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,3584,0.013786666923099093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,3584,0.030435555511050757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,2560,0.0169493340783649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,3584,0.035432888401879206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,3072,0.01274311128589842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,3072,0.027079999446868896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,2048,0.014139556222491793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,3072,0.03141866789923774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,2560,0.01165600038237042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,2560,0.023221333821614582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,1536,0.01143555591503779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,2560,0.028597331709331934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,2048,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,2048,0.01960355540116628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,1024,0.008471999731328752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,2048,0.024697777297761705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,10240,0.05265955461396111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,1536,0.009673777553770278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,1536,0.016754666964213055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,1024,0.008982222113344405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,1536,0.02036622166633606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,1024,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,5120,0.04604533314704895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,1024,0.017224889662530687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,768,0.011383111278216044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,256,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,768,0.015844444433848064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,512,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,512,0.014487110906177096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,128,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,256,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,256,0.012755555411179861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,128,0.007701333198282454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,64,0.003576000117593341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,3072,128,0.010668444136778513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,32,0.0038195554580953387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,768,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,64,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,32,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,3072,512,0.006065777606434292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,3072,768,0.008633777499198914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,65536,0.27318933275010854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,512,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,65536,0.14998932679494223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,51200,0.22995021608140734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,256,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,65536,0.42830755975511337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,16384,0.0786915553940667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,51200,0.11947111288706462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,12288,0.06776177883148193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,16384,0.038937777280807495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,51200,0.33842934502495664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,3072,128,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,16384,0.10258399777942234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,10240,0.05432266659206814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,12288,0.031277332040998675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,51200,0.34219198756747776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,12288,0.07881600326961942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,8192,0.04290133383538988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,10240,0.027680888772010803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,10240,0.06669600142372979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,10240,0.0786915553940667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,7168,0.034845332304636635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,8192,0.022780444886949327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,8192,0.05448266532686022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,8192,0.06422222322887845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,6144,0.03518133362134298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,7168,0.020807110601001315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,7168,0.04842133323351542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,7168,0.05867555406358507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,5120,0.028596444262398615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,6144,0.018616000811258953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,6144,0.042465776205062866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,65536,0.44147200054592556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,6144,0.051626665724648364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,4096,0.023183110687467787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,5120,0.016879111528396606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,5120,0.03708622190687392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,16384,0.11768978171878391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,5120,0.044927110274632774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,4096,0.014717333846622042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,4096,0.031142221556769475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,3072,0.019181332654423185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,12288,0.09027910894817776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,4096,0.03770400087038676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,3584,0.013632000320487551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,3584,0.027672888504134283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,2560,0.0162764439980189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,3584,0.03459822138150533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,3072,0.012514666550689273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,3072,0.024381332927280005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,2048,0.013462222284740873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,3072,0.03092799915207757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,2560,0.02122666603989071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,1536,0.010771555205186209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,2048,0.010625777973069085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,2048,0.018127111925019156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,2048,0.023697778582572937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,1024,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,1536,0.009661333428488838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,1536,0.015325332681337992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,1536,0.020238222347365487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,768,0.007370666497283512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,1024,0.009016888837019602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,3584,0.021290666527218286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,1024,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,512,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,1024,0.0169831116994222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,768,0.00848355558183458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,768,0.010573333336247338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,256,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,768,0.015504888362354703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,512,0.014089778065681458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,256,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,2560,0.011579555769761404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,256,0.00853066643079122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,256,0.01276000009642707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,2560,0.027236445082558527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,64,0.0038942222793896994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,128,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,32,0.004574222283230888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,128,0.00813600007030699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2560,128,0.010348444183667501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,64,0.007958222594526079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,65536,0.23680533303154838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,32,0.00812266684240765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,51200,0.17789688375261095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,65536,0.1424071126514011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,65536,0.372416893641154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,16384,0.07238577471839057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2560,512,0.009128889275921715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,51200,0.2906435595618354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,51200,0.11273688740200466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,65536,0.43122045199076336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,12288,0.05485333336724175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2560,128,0.004201777693298128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,16384,0.037548445992999606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,16384,0.09016178051630656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,10240,0.04398222102059258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,51200,0.3391315672132704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,12288,0.06901244322458903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,16384,0.11648178100585938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,8192,0.03408799899948968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,12288,0.0897208915816413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,10240,0.02683466672897339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,10240,0.05893599987030029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,10240,0.07685599724451701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,8192,0.021779555413458083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,8192,0.04818311002519396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,6144,0.024889777104059856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,7168,0.020049777295854356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,7168,0.04308799902598063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,7168,0.05741244554519653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,6144,0.018051554759343464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,5120,0.022653333014912073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2560,512,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,6144,0.03781422310405307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,6144,0.05031822125116984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,4096,0.01824977828396691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,5120,0.01590666671593984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,5120,0.03293066554599338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,3584,0.017176000608338248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,5120,0.043916443983713783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,4096,0.01421511173248291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,4096,0.02776266634464264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,4096,0.037263999382654824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,3072,0.01477777792347802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,3584,0.013420444395807056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,3584,0.024540444215138752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,7168,0.031446221801969744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,2560,0.01310222182008955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,3584,0.03394755721092224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,3072,0.012425777812798819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,3072,0.02200177808602651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,2048,0.011160888605647616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,8192,0.06488444407780965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,3072,0.030390222867329914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,2560,0.011413333316644033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,2560,0.01921422282854716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,1536,0.008647999829716152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,2560,0.026983999543719824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,2048,0.010324444207880232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,1024,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,1536,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,2048,0.023640000157886084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,1536,0.013788444300492605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,768,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,1536,0.020084444019529555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,1024,0.011124444504578909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,1024,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,512,0.005140444470776452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,12288,0.0302746660179562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,1024,0.016921778519948322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,768,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,768,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,256,0.0041866666740841335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,768,0.015459555718633862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,512,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,512,0.008520888785521189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,256,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,512,0.014105778601434497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,256,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,64,0.0032239999208185407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,256,0.012409778104888068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,32,0.0034808889031410217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,2048,128,0.0075395554304122925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,128,0.007344889144102733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,2048,128,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,64,0.007322666545708974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,32,0.007419555551475949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,51200,0.13719111018710667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,65536,0.09710577461454604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,65536,0.28669510947333443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,16384,0.053456889258490674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,51200,0.07742755280600654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,65536,0.29763110478719074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,51200,0.22670221328735352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,2048,128,0.003546666767862108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,16384,0.07198666863971286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,12288,0.03990044527583652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,16384,0.026185777452256944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,51200,0.2317333221435547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,10240,0.03319022059440613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,16384,0.08220089144176908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,12288,0.021454221672481958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,8192,0.02584533393383026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,12288,0.0549759997261895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,10240,0.019651555352740817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,12288,0.06466133064693876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,7168,0.02204888893498315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,10240,0.05542666713396708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,2048,2048,0.016550223032633465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,8192,0.015847111741701763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,8192,0.03905511233541701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,8192,0.04628355635537041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,7168,0.014492443866199918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,7168,0.03492533498340183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,5120,0.016247999336984422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,7168,0.04186933239301046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,6144,0.013168888787428537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,6144,0.03103644318050808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,4096,0.013720889058378009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,5120,0.011875555747085147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,5120,0.026755554808510676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,5120,0.03254310952292548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,4096,0.02204977803760105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,3584,0.012297777665985955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,4096,0.011044444309340583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,4096,0.028239998552534316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,3072,0.011002666420406766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,3584,0.010400888820489248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,3584,0.019882667395803664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,3584,0.025091555383470323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,2560,0.009559111462699043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,65536,0.21677066220177543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,3072,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,3072,0.018198221921920776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,6144,0.020583111378881667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,3072,0.023016888234350417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,2560,0.009376000199053023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,2560,0.015824000040690105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,1536,0.0070168889231152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,2048,0.008952889177534316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,2048,0.013495999905798169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,2048,0.018544000056054857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,6144,0.0370399985048506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,1024,0.00534400012758043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,1536,0.011736888852384357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,1536,0.00850577818022834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,1536,0.016149333781666227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,768,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,1024,0.007874666816658443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,1024,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,1024,0.014130666851997375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,512,0.0041128889554076726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,768,0.00835111074977451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,768,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,512,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,2048,0.008286221987671321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,256,0.0033075554917256036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,512,0.012295111185974546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,512,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,128,0.0029084444459941653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,2560,0.02087200019094679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,256,0.006923555499977536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,256,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,64,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,256,0.011077333655622272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,128,0.007110222346252865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1536,32,0.0028435554769304064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1536,128,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,64,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,32,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1536,10240,0.04692711101637947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,768,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,65536,0.09228889147440593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,65536,0.22915644115871855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,51200,0.10899288786782159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,16384,0.039579557047949895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,65536,0.28973423110114205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,51200,0.07424533367156982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,51200,0.1822293334537082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,12288,0.02879377868440416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,16384,0.025030222203996446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,51200,0.226127995385064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1536,128,0.0069013333155049225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,16384,0.0802275538444519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,12288,0.020650666620996263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,12288,0.045648889409171216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,8192,0.02027466727627648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,65536,0.15889067120022243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,12288,0.06292088826497395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,10240,0.018909333480728995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,10240,0.03894044293297662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,7168,0.01777333352300856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,8192,0.01520088811715444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,8192,0.03278933299912347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,6144,0.015431111057599386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,8192,0.0450799994998508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,7168,0.01404622197151184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,7168,0.02940088841650221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,5120,0.01344622257683012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,7168,0.04090044564670987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,6144,0.012784000072214337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,6144,0.025888888372315302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,4096,0.011083555718262991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,5120,0.011771555576059552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,6144,0.035805334647496544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,16384,0.05964177846908569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,3584,0.010395555860466426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,4096,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,4096,0.0188426673412323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,10240,0.024131556351979572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,3072,0.009129777550697327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,4096,0.026504889130592346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,3584,0.010068444742096795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,3584,0.02444800072246128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,3072,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,3072,0.015511999527613321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,10240,0.05386666787995232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,3072,0.02218399941921234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,2048,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,2560,0.009098666409651438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,2560,0.013850666582584381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,2560,0.020263110597928364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,1536,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,2048,0.012061333490742577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,2048,0.01791111131509145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,5120,0.02204799983236525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,1024,0.004855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,1536,0.01036711119943195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,5120,0.031275557147132024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,1536,0.00812711069981257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,1536,0.01551644504070282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,768,0.004208000169859992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,1024,0.007627555893527136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,1024,0.008618666893906063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,3584,0.017136888371573556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,512,0.0035031110876136353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,1024,0.013780444032616086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,768,0.007592889169851939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,256,0.0032088889016045462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,512,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,512,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,512,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,256,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,256,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,128,0.00315022220214208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,256,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,2048,0.008658666577604082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,64,0.0028782222006056043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,128,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,128,0.005975111077229182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,32,0.002874666617976295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,128,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,64,0.006049777898523543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,1024,32,0.006215999937719769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,1024,768,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,65536,0.1881706714630127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,65536,0.09135999944474961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,1024,768,0.01274311128589842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,65536,0.2781386640336779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,51200,0.07357333103815715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,16384,0.03106844425201416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,51200,0.14914222558339438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,51200,0.2191173368030124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,16384,0.024616888827747766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,12288,0.02021511064635383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,16384,0.04954044355286492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,10240,0.016972444123691983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,16384,0.07743200328614977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,12288,0.020086222224765353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,65536,0.125453339682685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,8192,0.01497244503762987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,12288,0.060431996981302895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,10240,0.018541332748201158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,10240,0.03312266535229153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,7168,0.012816888590653738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,10240,0.05224711034033033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,8192,0.014831999937693277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,8192,0.027592000034120347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,1024,2560,0.008021333151393467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,6144,0.011511999699804517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,7168,0.013810666898886362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,7168,0.025175111161337957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,5120,0.010093332992659675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,7168,0.039091553952958845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,6144,0.012784000072214337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,6144,0.022336888644430373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,4096,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,6144,0.03420977791150411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,5120,0.011726222104496427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,5120,0.018915555543369718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,3584,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,5120,0.030144890149434406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,4096,0.0107342220014996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,4096,0.016182222300105624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,3072,0.007233777807818518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,4096,0.02567200031545427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,3584,0.010178666975763109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,12288,0.03876266545719571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,3584,0.014825777875052558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,2560,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,3584,0.023725334140989516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,3072,0.009662222531106737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,3072,0.01350044459104538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,2048,0.005428444594144821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,2560,0.009176889227496253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,3072,0.02161866592036353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,2560,0.012085333466529846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,1536,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,51200,0.09539110793007745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,2560,0.019668444991111755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,8192,0.04366844561364916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,2048,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,2048,0.01757066614098019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,1536,0.008033778104517195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,1536,0.009317333499590555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,768,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,1536,0.015444444285498725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,1024,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,512,0.003024000053604444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,1024,0.01350311107105679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,768,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,768,0.012786666552225748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,256,0.0026951111439201566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,512,0.006330666856633292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,512,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,512,0.01179377817445331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,256,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,128,0.0025235555238193935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,256,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,256,0.010688888529936472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,64,0.0022088889446523455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,128,0.006653333289755716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,128,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,768,128,0.008660444782839881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,2048,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,65536,0.08910399675369263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,64,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,32,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,768,1024,0.00795555528667238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,51200,0.06542933649486966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,65536,0.09120800097783406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,768,768,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,65536,0.18153511153327095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,16384,0.024460444847742718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,51200,0.07351022296481662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,51200,0.14335821734534368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,16384,0.04825422167778015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,51200,0.21667022175259062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,16384,0.0240657776594162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,12288,0.02748888896571265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,16384,0.07630222373538546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,12288,0.019683554768562317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,32,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,10240,0.022656889425383672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,12288,0.03782577647103204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,768,1024,0.003830222205983268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,8192,0.019163555569118924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,12288,0.059584889147016734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,10240,0.01824355622132619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,7168,0.016807110773192514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,8192,0.014483556151390076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,10240,0.051388445827696055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,8192,0.026848889059490625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,6144,0.014774221513006421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,8192,0.04305244485537211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,7168,0.01348888874053955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,7168,0.024403555525673762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,5120,0.01240533341964086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,7168,0.038110223081376814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,6144,0.012409778104888068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,6144,0.021496888664033677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,4096,0.008077333370844523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,6144,0.03344355689154731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,5120,0.01142133358452055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,5120,0.018792889184421964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,3584,0.007463999920421177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,4096,0.01036444471942054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,5120,0.029350221157073975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,4096,0.0159262220064799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,3072,0.006872889068391588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,4096,0.025064888927671645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,3584,0.009768889182143742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,3584,0.014493332968817817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,2560,0.006022222340106964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,3072,0.009409777820110321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,3584,0.023375999596383836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,3072,0.013224889006879596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,2048,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,10240,0.03249333302179972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,3072,0.021024000313546922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,2560,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,1536,0.004599110947714912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,2560,0.019497777024904888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,2048,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,2048,0.010382222632567087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,2048,0.01724799970785777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,1536,0.00795555528667238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,1536,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,768,0.0031893334041039148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,1536,0.01496355566713545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,65536,0.2733164363437229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,1024,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,1024,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,512,0.0028151110228565005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,768,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,768,0.01181422256761127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,256,0.0025333333760499954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,512,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,512,0.01142133358452055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,256,0.005684444473849402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,256,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,128,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,256,0.01036444471942054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,2560,0.011752888560295105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,64,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,128,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,1024,0.0038853333228164246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,128,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,512,32,0.0024871110088295406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,128,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,65536,0.04898577928543091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,64,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,51200,0.041327999697791204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,32,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,65536,0.09091111024220784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,65536,0.14169511530134413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,512,1024,0.01311555587583118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,51200,0.07333422369427152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,512,512,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,16384,0.03822311096721225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,16384,0.02329155637158288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,12288,0.013201778133710226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,16384,0.07500355773501925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,12288,0.02997244397799174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,12288,0.01957066688272688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,10240,0.012096000214417776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,12288,0.05843466520309448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,10240,0.018107554978794523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,8192,0.012416889270146688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,10240,0.02570044497648875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,10240,0.05018844538264804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,512,768,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,8192,0.021680000755521987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,7168,0.011047110789351992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,8192,0.04191199938456217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,8192,0.014375999569892883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,51200,0.21451111634572348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,7168,0.013407111167907715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,7168,0.01960355540116628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,7168,0.03714044557677375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,6144,0.01754666699303521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,5120,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,6144,0.012330667012267642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,6144,0.032814221249686346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,5120,0.015495111544926962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,4096,0.007257777783605788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,5120,0.028775112496482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,16384,0.016414221790101793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,65536,0.2702702151404487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,4096,0.013452444639470843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,3584,0.006650666809744305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,4096,0.0248355550898446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,6144,0.010021333065297868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,3072,0.006690666493442323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,3584,0.009694221946928237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,3584,0.022875555687480505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,3072,0.011410666836632622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,3072,0.009322666459613377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,2560,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,3072,0.02081155611409081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,2560,0.008976889153321585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,5120,0.011392888923486074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,2048,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,2560,0.0103502223889033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,2560,0.018943111101786297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,1536,0.0041999999019834726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,2048,0.008375110725561777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,2048,0.009396444592210982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,51200,0.1110337773958842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,4096,0.01018488903840383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,1024,0.0035217776894569397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,2048,0.016866667403115165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,1536,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,1536,0.008272888759771982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,1024,0.007495111061467065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,1024,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,768,0.006315555423498154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,768,0.011760888828171624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,512,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,512,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,512,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,256,0.002500444443689452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,256,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,256,0.009848888549539778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,128,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,256,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,768,0.003142222141226133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,1536,0.014485332700941296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,64,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,128,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,3584,0.012373333175977072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,32,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,128,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,65536,0.0396106673611535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,128,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,64,0.005047111047638787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,51200,0.031848000155554876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,256,32,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,65536,0.09068089061313206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,65536,0.1403048833211263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,16384,0.012120889292822944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,51200,0.07290044758054945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,51200,0.1105075544781155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,65536,0.27028266588846844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,12288,0.010269333091047075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,51200,0.21372000376383463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,16384,0.022848889231681824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,16384,0.07543910874260797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,12288,0.029504888587527808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,10240,0.009356444080670675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,12288,0.05812178055445353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,256,1024,0.013122667041089801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,8192,0.008032889001899296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,10240,0.025240888198216755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,10240,0.0177208897140291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,10240,0.04965510964393616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,7168,0.007698666718271043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,8192,0.014422222971916199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,8192,0.021317332983016968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,8192,0.04119288921356201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,6144,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,7168,0.013440888788965015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,7168,0.03688800003793504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,16384,0.037449777126312256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,5120,0.006719111154476802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,6144,0.017465778523021273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,6144,0.012366222010718452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,6144,0.03261955579121908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,12288,0.019289778338538278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,4096,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,5120,0.011214222345087262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,5120,0.015207999282413058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,5120,0.028725332683987085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,256,512,0.0028151110228565005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,3584,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,4096,0.013350221845838757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,4096,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,3072,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,4096,0.0247653325398763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,3584,0.00978311151266098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,3584,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,2560,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,3584,0.02275466587808397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,3072,0.009305777649084726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,3072,0.01127644462717904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,2048,0.004573333180612988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,3072,0.020844444632530212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,7168,0.01926577753490872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,2560,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,2560,0.010330666270520952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,1536,0.003900444342030419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,2048,0.008548444343937768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,2560,0.01887022289964888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,2048,0.00904266701804267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,1024,0.003177777760558658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,2048,0.016872000363137987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,1536,0.008278222547637092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,768,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,1536,0.01443288889196184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,1024,0.007330666813585494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,1024,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,512,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,1024,0.01277955538696713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,256,768,0.007196444604131911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,768,0.007058666812049017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,256,0.002492444382773505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,768,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,512,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,768,0.011697777443461947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,128,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,512,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,512,0.01106844428512785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,256,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,64,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,256,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,128,32,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,256,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,128,0.006593777901596493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,128,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,65536,0.037590222226248846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,51200,0.0318488876024882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1536,128,128,0.007808888951937358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,16384,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,12288,0.010001777774757808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,32,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,10240,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,65536,0.14147999551561144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1536,128,1536,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,51200,0.11019644472334121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,16384,0.037010666396882795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,8192,0.00816888858874639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,7168,0.007874666816658443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,12288,0.029162665208180744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,6144,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,5120,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,4096,0.0064488889442549805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,10240,0.025267556309700012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,3584,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,3072,0.005535111245181825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,8192,0.02130311065249973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,7168,0.019239111079110038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,2560,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,5120,0.015293333265516492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,2048,0.004555555681387584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,4096,0.01314311143424776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,1536,0.0038844446341196695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,1024,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,3584,0.012446222205956778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,768,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,2560,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,512,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,2048,0.008984000318580205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,1536,0.007968000239796108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,1024,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,128,0.0021911110315057966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,768,0.006305777778228124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,64,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,512,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,128,64,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,32,0.0019466665883858998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,256,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,65536,0.03570488757557339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,128,0.005142222262091107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,51200,0.02808622188038296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,16384,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,32,0.004961777892377642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,12288,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,65536,0.14055644141303167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,51200,0.11007644732793172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,10240,0.014787554740905762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,16384,0.037059555451075234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,6144,0.017500445246696472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,12288,0.028885334730148315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,7168,0.012054222325483957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,3072,0.011358222199810876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,10240,0.02514933380815718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,6144,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,8192,0.02130844526820713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,5120,0.009810667071077559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,64,256,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,7168,0.019271110494931538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,3584,0.00886044485701455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,6144,0.017553778158293832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,3072,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,5120,0.015472888946533203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,2560,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,4096,0.013415111435784234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,2048,0.005046222358942032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,3584,0.012284444438086616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,64,64,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,3072,0.011337777806652917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,1024,0.00315022220214208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,2560,0.010110222631030613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,768,0.002789333255754577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,2048,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,512,0.002487999904486868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,1536,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,8192,0.012853333519564735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,1024,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,128,0.001826666709449556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,768,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,64,0.0018560000591807896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,256,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,32,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,128,0.005008888741334279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,4096,0.008912888665994009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,64,0.005048000150256687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,32,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,1536,0.0035582222044467926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1536,32,256,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1536,32,512,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,16384,0.5792097515530056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,16384,1.752253320482042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,12288,1.3678230709499781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,16384,1.8575564490424261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,12288,0.44523021909925675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,12288,1.318177753024631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,10240,0.37764443291558164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,10240,1.0419119728936088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,51200,2.0192445119222007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,8192,0.9276640150282117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,10240,1.0963786443074544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,8192,0.30431556701660156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,8192,0.8572506904602051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,7168,0.7023244433932834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,7168,0.32100534439086914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,6144,0.24409333864847818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,6144,0.6148097780015733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,5120,0.505715529123942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,7168,0.8964106241861979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,6144,0.6694755554199219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,5120,0.207677337858412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,4096,0.43753332561916775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,5120,0.5723724365234375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,4096,0.16747732957204184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,3584,0.36189688576592344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,4096,0.43723734219868976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,3072,0.2807626724243164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,3584,0.1548666689130995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,3072,0.14835555023617214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,3584,0.3773342238532172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,2560,0.24010311232672799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,3072,0.3221280044979519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,51200,5.541886223687066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,2560,0.11287377940283881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,2048,0.20958222283257377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,1536,0.16845333576202393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,2048,0.09794133239322239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,2560,0.2815351221296522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,1024,0.11248266696929932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,1536,0.07588355408774482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,2048,0.22223022249009874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,768,0.10155200295978123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,1024,0.061758220195770264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,512,0.07641510830985175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,1024,0.13307555516560873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,768,0.05576444334454007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,768,0.10936888721254136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,256,0.0437155564626058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,512,0.05151289039187961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,128,0.04059111078580221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,512,0.09182933304044937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,1536,0.17827288309733072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,64,0.035599999957614474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,65536,32,0.039148443275027804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,256,0.08291377623875935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,128,0.04541511005825467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,51200,6.425718943277995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,128,0.07752355602052477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,64,0.07756266991297404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,65536,32,0.07805955410003662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,65536,256,0.047232889466815524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,16384,1.353282716539171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,51200,1.6276026831732855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,65536,2.085478252834744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,16384,0.4473590850830078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,16384,1.5228222741021051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,12288,1.187345822652181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,12288,0.34139466285705566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,51200,4.165072123209636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,10240,0.8027644687228732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,10240,0.28648445341322154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,12288,1.1747155719333224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,65536,5.210914611816406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,8192,0.23633954260084364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,8192,0.5795715649922689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,51200,5.023403591579862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,10240,0.938776863945855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,8192,0.7251271141899956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,7168,0.20835199620988634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,6144,0.43160443835788304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,6144,0.18466133541531035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,5120,0.3710373242696126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,6144,0.5131751166449653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,5120,0.1652639971839057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,4096,0.35416444142659503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,7168,0.5878435770670573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,5120,0.4513111114501953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,4096,0.13568354977501765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,3584,0.2751235432094998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,3584,0.11730044417911106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,3072,0.2679813437991672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,3584,0.35328443845113117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,3072,0.10373866558074951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,2560,0.2232444551255968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,7168,0.6127449141608344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,3072,0.24923822614881727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,2048,0.17243556181589761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,2560,0.09054133627149795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,65536,7.189114888509114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,2048,0.07800177733103435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,1024,0.09617333279715644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,1536,0.13391555680169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,2048,0.19843467076619467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,1536,0.060770663950178355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,768,0.08119199673334758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,1536,0.13974755340152317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,1024,0.052110221650865346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,512,0.05673066775004069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,768,0.04559288753403557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,1024,0.10487022002538045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,256,0.037244445747799344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,128,0.031022220849990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,512,0.042275554604000516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,768,0.08672622177335952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,4096,0.35062045521206325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,256,0.03839822279082404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,512,0.07309866613811918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,64,0.031389332479900785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,51200,32,0.031017776992585924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,51200,128,0.03773422373665704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,256,0.06604977448781331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,128,0.06313333246443006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,64,0.06193777587678698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,32,0.06218577755822075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,51200,2560,0.23565867212083605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,16384,0.38472265667385525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,65536,0.6673626899719238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,51200,0.5231128798590766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,16384,0.15819022390577528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,51200,1.5344764921400282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,12288,0.386519988377889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,65536,1.7487359576755102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,12288,0.12168355782826741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,12288,0.38446932368808323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,51200,1.575201776292589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,10240,0.317138671875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,8192,0.2080355617735121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,10240,0.10264977481630112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,10240,0.3168124357859294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,7168,0.21093332767486572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,8192,0.08578488561842178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,65536,2.207096947564019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,8192,0.22655110889010957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,7168,0.07424533367156982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,6144,0.1785564422607422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,7168,0.19374577204386392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,16384,0.4469173219468858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,6144,0.06675022178226046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,5120,0.1561839977900187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,4096,0.1283528937233819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,5120,0.05602577659818861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,3584,0.11125777827368842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,4096,0.046662221352259316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,5120,0.14093689123789468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,6144,0.17152622010972765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,3072,0.09444799688127305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,3584,0.04303288790914747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,2560,0.07597955730226305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,3584,0.09906578063964844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,2048,0.06825244426727295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,3072,0.037696888049443565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,3072,0.09340266386667888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,1536,0.046839111381106906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,2560,0.07570222351286146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,1024,0.032736887534459434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,2048,0.06171822547912598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,1536,0.02381155557102627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,768,0.025830222500695124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,1536,0.049125333627065025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,1024,0.019865777757432725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,512,0.019455111689037748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,1024,0.03791022300720215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,768,0.018051554759343464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,256,0.014299554957283868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,768,0.03137244449721442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,512,0.025504888759719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,256,0.015152888165579902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,128,0.011764444410800934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,256,0.023120888405376013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,64,0.01036533299419615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,128,0.015248888068728976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,4096,0.11273954974280463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,16384,32,0.01037333326207267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,128,0.022035555707083807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,2560,0.03393155667516921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,2048,0.028705779049131606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,64,0.022042666872342426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,16384,32,0.022057778305477563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,16384,512,0.01645511057641771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,65536,0.4829297595553928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,16384,0.27541155285305446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,51200,0.38471200731065536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,51200,0.9207848442925347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,12288,0.2059475580851237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,16384,0.14281866285536024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,65536,1.3130142423841689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,16384,0.34350132942199707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,12288,0.09469778007931179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,10240,0.0937831136915419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,8192,0.14305422041151258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,51200,1.2431324852837455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,7168,0.12131733364529079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,65536,1.5805928972032335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,10240,0.2926284472147624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,8192,0.06774577829572889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,6144,0.1189075575934516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,8192,0.17144266764322916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,7168,0.056780444251166455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,7168,0.14730755488077799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,5120,0.09234044286939834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,6144,0.054656889703538686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,4096,0.0903351108233134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,6144,0.13117155763838026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,5120,0.04639377858903673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,3584,0.07098666826883952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,4096,0.03649955656793382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,3072,0.05646755297978719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,5120,0.10960266987482707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,3584,0.035247110658221774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,4096,0.08850844701131184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,2560,0.05577066871854994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,3584,0.07820622126261394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,3072,0.02940711047914293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,2048,0.04434400134616428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,2560,0.026900443765852187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,1536,0.032532443602879844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,3072,0.06808533271153767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,10240,0.18635555108388266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,2560,0.05917688873079088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,2048,0.04900977677769131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,768,0.018251554833518136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,1024,0.016536888149049547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,1536,0.03924533393647935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,512,0.013789333403110504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,1024,0.029934220843844946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,768,0.015114666687117683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,768,0.025123554799291823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,512,0.013735111388895245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,512,0.020672889219390023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,128,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,256,0.012708444562223224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,256,0.01825777855184343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,64,0.007672000262472365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,128,0.013042666845851474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,32,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,128,0.017887999614079792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,64,0.017811555001470778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,2048,0.02290222214327918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,12288,1536,0.019489778412712943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,256,0.010538666612572141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,65536,0.439716445075141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,65536,1.0452719794379342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,51200,0.7759831216600207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,16384,0.2839511235555013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,12288,0.2821528911590576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,51200,0.3399191167619493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,65536,1.403958214653863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,16384,0.31202756034003365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,16384,0.1129608949025472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,12288,0.18811110655466715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,51200,0.9792977439032661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,12288,0.08585688802931045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,12288,0.22806223233540854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,10240,0.06803466876347859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,10240,0.1835191117392646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,7168,0.1430951091978285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,8192,0.05504711137877571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,8192,0.1416648891237047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,7168,0.04917777909172905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,6144,0.12199644247690837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,7168,0.12553332911597356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,12288,32,0.017656000124083627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,5120,0.10446488857269287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,6144,0.043508443567487926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,6144,0.10943288935555352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,4096,0.0833502213160197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,5120,0.03734666771358914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,5120,0.09201155768500434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,10240,0.16431821717156306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,3584,0.07257066832648383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,4096,0.03149688906139798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,3072,0.061695999569363065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,4096,0.07500266366534762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,8192,0.16430400477515325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,3584,0.028427554501427546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,2560,0.04368533359633552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,3584,0.06577510966195001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,2048,0.036269333627488874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,2560,0.023373333944214716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,3072,0.057215107811821826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,1536,0.02955733405219184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,2560,0.05048977666431003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,2048,0.02079288827048408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,1024,0.02055111030737559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,1536,0.01763199932045407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,2048,0.04167377617624071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,768,0.016193777322769165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,1536,0.03388799892531501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,1024,0.014886221951908536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,512,0.012803555362754397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,12288,1024,0.024244444237815008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,1024,0.026200888885392085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,256,0.009151111046473185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,512,0.012658666405412885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,768,0.021872889664438035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,256,0.011728000309732227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,512,0.018242667118708294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,64,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,32,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,256,0.01614488826857673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,128,0.011791110866599612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,128,0.01555377741654714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,64,0.01575999955336253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,10240,32,0.01554844445652432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,3072,0.026183999247021143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,65536,0.3572479883829753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,65536,0.8185120158725314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,10240,768,0.013714666995737286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,16384,0.23993155691358778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,51200,0.7834329075283475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,10240,128,0.00793599999613232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,16384,0.08876177999708389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,51200,0.7810986836751302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,12288,0.1896746688418918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,12288,0.06820799907048543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,16384,0.2571688758002387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,10240,0.16262488894992405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,12288,0.19632089138031006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,10240,0.07150399684906006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,8192,0.047520889176262744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,7168,0.11813244554731582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,10240,0.14495021767086452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,6144,0.0947164429558648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,8192,0.12299377388424343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,7168,0.042711999681260854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,5120,0.07578844494289823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,7168,0.129856891102261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,6144,0.0389066669676039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,6144,0.08993777963850234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,4096,0.07099022467931111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,51200,0.2728346718682183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,5120,0.03286844491958618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,3584,0.05832799937989977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,5120,0.0762959983613756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,4096,0.02783999840418498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,3072,0.050824890534083046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,3584,0.02530133393075731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,4096,0.06245511107974582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,3584,0.0554711123307546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,65536,1.0979422463311088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,2560,0.021000888612535264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,3072,0.04845955636766222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,2560,0.04246844516860115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,1536,0.02753866712252299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,2048,0.018703111343913607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,8192,0.1349617772632175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,1024,0.01698666645420922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,2048,0.03545422355333964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,1536,0.01579466627703773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,768,0.013594666288958656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,1024,0.013400000002649097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,1024,0.02238311039076911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,512,0.010980444649855295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,768,0.018599111172888014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,512,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,256,0.008241777618726095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,512,0.015440000428093804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,2560,0.04364533225695292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,256,0.010728889041476779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,3072,0.023178666830062866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,256,0.013927999469969006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,2048,0.03597955571280585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,64,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,32,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,128,0.01074399964676963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,128,0.013439111411571503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,64,0.013437333206335703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,32,0.013447110851605734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,8192,1536,0.028990222348107234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,8192,768,0.012392889294359418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,65536,0.3069795502556695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,65536,0.7547155486212836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,51200,0.5883377922905816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,51200,0.24212800131903756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,8192,128,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,16384,0.07904177904129028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,65536,0.9518444273206922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,12288,0.16944355434841582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,10240,0.13772000206841364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,12288,0.06082666582531399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,12288,0.15404266781277126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,8192,0.09731466902626885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,10240,0.054692443874147206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,10240,0.1287955575519138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,8192,0.0413262214925554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,7168,0.09613777531517877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,8192,0.10703555742899577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,7168,0.03775022096104092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,6144,0.0813831090927124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,5120,0.06820178031921387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,6144,0.03293600016170078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,6144,0.08031377527448866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,4096,0.0572684473461575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,5120,0.029398222764333088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,5120,0.068159110016293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,16384,0.2151475614971585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,4096,0.024614221519894067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,4096,0.05525066786342197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,3072,0.04131999942991469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,3584,0.022662222385406494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,3584,0.049119111564424306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,2560,0.034993777672449745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,16384,0.23147556516859266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,3072,0.020767110917303298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,2048,0.02722666660944621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,3072,0.043178667624791466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,1536,0.02110044492615594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,2560,0.03776889046033224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,2048,0.016730666160583496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,2048,0.032187554571363665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,1024,0.01605599953068627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,1536,0.014496889379289416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,768,0.013169777890046438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,1536,0.025944888591766357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,1024,0.012686221963829465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,512,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,7168,0.09390488598081802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,1024,0.019744000501102872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,256,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,512,0.010841777755154504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,768,0.016922665966881644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,128,0.006392000036107169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,3584,0.04840266704559326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,256,0.010074666804737514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,256,0.012788444757461548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,64,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,128,0.010038221875826517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,51200,0.6962738037109375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,7168,32,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,128,0.012376888758606382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,64,0.012401777837011548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,2560,0.019075555933846366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,32,0.012387555506494312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,65536,0.5515368779500326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,65536,0.2626355489095052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,51200,0.4212426609463162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,51200,0.20522488488091362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,16384,0.1423057715098063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,7168,768,0.011689778003427716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,12288,0.12011288272009955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,65536,0.7782044410705566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,16384,0.06590933269924588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,10240,0.10717866818110149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,7168,512,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,51200,0.5982009039984809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,12288,0.051334222157796226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,8192,0.08453244633144802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,10240,0.044723557101355664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,8192,0.03666844301753574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,10240,0.11682844161987305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,6144,0.06226844257778592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,7168,0.032124443186653986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,8192,0.09180266989601983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,7168,0.09025155835681492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,5120,0.045194665590922035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,6144,0.028384000062942505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,4096,0.04495999879307217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,5120,0.025274667474958632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,6144,0.07116533650292291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,3584,0.03755555550257365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,5120,0.06122933493720161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,4096,0.02176355487770504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,3072,0.030902223454581365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,4096,0.049118220806121826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,3584,0.020353777541054618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,2560,0.02864177690611945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,3584,0.044234666559431285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,3072,0.018532445033391316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,2048,0.02271288964483473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,16384,0.183935112423367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,3072,0.03871377640300327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,2560,0.016948444975747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,1536,0.018588445252842374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,2560,0.03410755594571432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,12288,0.15182311005062527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,1024,0.014093332820468478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,2048,0.014967999524540372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,1536,0.013423999978436364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,2048,0.028760890165964764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,1024,0.011304000185595619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,512,0.00868711123863856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,1024,0.01777777738041348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,768,0.015453333655993143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,256,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,512,0.00999466660949919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,512,0.013023110727469126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,128,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,256,0.009319110876984065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,256,0.012113778127564324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,128,0.009336888790130615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,64,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,32,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,128,0.011353777514563667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,64,0.011391110718250275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,7168,0.07558311356438531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,32,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,6144,768,0.01254666679435306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,6144,1536,0.023608888188997906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,65536,0.5822630988226997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,6144,768,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,51200,0.3469359874725342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,16384,0.1351377831565009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,65536,0.2434577676984999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,12288,0.09115733040703668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,16384,0.06147022379769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,51200,0.18922044171227348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,10240,0.07714133130179511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,12288,0.05191555619239807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,8192,0.061717331409454346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,65536,0.6507004631890191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,51200,0.5126720004611546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,7168,0.06624444325764973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,10240,0.09718222088283962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,8192,0.033942222595214844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,6144,0.058989332781897656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,8192,0.08270577589670818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,7168,0.03195200032658047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,5120,0.04186222288343641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,7168,0.0705359975496928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,6144,0.027388445205158655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,4096,0.035164445638656616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,5120,0.024422221713595923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,6144,0.061887999375661217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,3584,0.030576000610987347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,5120,0.054465777344173856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,4096,0.020925333102544148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,3072,0.028652442826165095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,4096,0.04300355580117968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,3584,0.019645333290100098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,2560,0.024313777685165405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,3584,0.0388648874229855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,3072,0.017708443933063082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,2048,0.018367111682891846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,3072,0.034501334031422935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,1536,0.01588622232278188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,2560,0.030046221282747056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,2048,0.01440533333354526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,16384,0.20847022533416748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,1024,0.011602666642930774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,2048,0.025411556164423626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,1536,0.012807110945383707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,1536,0.021011556188265484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,1024,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,512,0.007171555525726742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,1024,0.016024000114864774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,768,0.010321777727868822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,256,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,768,0.013947554760509066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,512,0.009725333087974125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,128,0.005200888961553574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,256,0.009111999637550777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,256,0.010681777364677854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,128,0.009116444322797988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,128,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,64,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,32,0.004844444493452708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,64,0.01035822182893753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,12288,0.11487021711137559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,10240,0.04309599929385715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,65536,0.19412622186872694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,512,0.011786667009194693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,65536,0.5426586469014486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,51200,0.29818399747212726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,16384,0.08940533134672378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,5120,2560,0.01590488851070404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,51200,0.15100267198350695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,51200,0.4259502092997233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,16384,0.05016622278425428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,5120,32,0.01034577770365609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,16384,0.1296959982977973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,10240,0.08053777615229289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,12288,0.03979911075698005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,12288,0.09693777561187744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,10240,0.03495999839570787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,65536,0.4597955809699164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,10240,0.08220622274610731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,7168,0.05475022064314949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,8192,0.028227554427252874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,8192,0.06723911232418485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,7168,0.02563200063175625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,5120,0.039164443810780845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,7168,0.05906933546066284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,6144,0.023360000716315374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,4096,0.0307626658015781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,6144,0.05180444320042928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,5120,0.021145777569876775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,12288,0.08124888605541654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,3584,0.025063110722435847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,5120,0.04437866806983948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,4096,0.01816177699300978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,3072,0.02402666707833608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,5120,768,0.009062222308582729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,4096,0.03735911183887058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,2560,0.020063110523753695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,3584,0.016588444511095684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,3072,0.01536355581548479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,2048,0.01648355523745219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,3584,0.03384266628159417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,8192,0.06616355313195123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,1536,0.013385777672131857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,3072,0.03025955624050564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,1024,0.010312000082598792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,2560,0.026199110680156287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,2048,0.012668444050682915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,768,0.008271111382378472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,1536,0.011217777927716574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,2048,0.02211644417709774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,512,0.006476444502671559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,1024,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,1536,0.018248889181349013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,6144,0.04801511102252536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,768,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,512,0.008737777670224508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,768,0.01238666640387641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,128,0.004252444538805219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,256,0.008375110725561777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,512,0.010464888479974536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,64,0.003891555385457145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,256,0.00962577760219574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,128,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,128,0.009305777649084726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,64,0.009308444129096137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,32,0.009302222066455418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,65536,0.4362391100989447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,4096,2560,0.013735999663670858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,65536,0.1867084503173828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,51200,0.2811422348022461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,16384,0.11200888951619466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,65536,0.478765328725179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,51200,0.14655200640360513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,256,0.005190222213665645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,4096,1024,0.014167999227841696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,12288,0.069504890176985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,16384,0.04888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,16384,0.11494133207533096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,10240,0.059602664576636426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,8192,0.0513937771320343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,51200,0.43084534009297687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,10240,0.03379111157523261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,7168,0.04566311173968845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,4096,32,0.003926222109132343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,12288,0.08567733234829372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,8192,0.02751200066672431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,8192,0.05996711386574639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,7168,0.026421333352724712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,5120,0.03236711025238037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,6144,0.023190221852726404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,4096,0.027759111589855615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,6144,0.04619910981920031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,5120,0.02030755579471588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,5120,0.040157334672080144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,4096,0.017535110314687092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,4096,0.03344444433848063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,3584,0.01601777805222405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,3584,0.02985066506597731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,12288,0.038872000243928694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,3072,0.026369776990678575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,2560,0.017671111557218764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,3072,0.014765333798196582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,2560,0.013520888984203339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,7168,0.05266133281919691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,2048,0.015249777171346875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,2560,0.02310666607485877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,1536,0.012135999898115793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,2048,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,3584,0.02272355556488037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,2048,0.019711110326978896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,1024,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,1536,0.011043555206722684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,1536,0.01611022154490153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,768,0.007845333053006066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,3072,0.021521778570281133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,1024,0.009729777773221334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,1024,0.012770666844315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,768,0.00909511082702213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,256,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,768,0.011413333316644033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,512,0.00869511150651508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,128,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,256,0.008270222279760573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,512,0.009725333087974125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,10240,0.07290222247441609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,256,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,32,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,128,0.008722667064931657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,64,0.008637333081828224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3584,32,0.008646222452322641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,65536,0.3596666653951009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,51200,0.23153334193759492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,65536,0.15147999922434488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,16384,0.08779644303851658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,512,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,65536,0.42053598827785915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,51200,0.12391111585828994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,12288,0.05932088692982992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,16384,0.039878222677442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,10240,0.05587999688254463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,16384,0.10170311397976345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,6144,0.039375109805001154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,12288,0.032460444503360324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3584,128,0.00835111074977451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,12288,0.07818044556511773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,10240,0.028118222951889038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,8192,0.022996443841192458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,10240,0.065647111998664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,6144,0.03414222266938951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,8192,0.05420800050099691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,7168,0.020696888367335003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,7168,0.048026667700873486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,5120,0.028003556860817805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,6144,0.018918222851223417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,4096,0.02254933284388648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,6144,0.0422053337097168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,5120,0.016677333248986136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,3584,0.020026667250527274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,5120,0.03608355588383145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,4096,0.014456889695591398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,3072,0.018062222335073683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,4096,0.030321776866912842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,3584,0.01349422252840466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,51200,0.39748533566792804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,2560,0.01443377799457974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,3584,0.027095999982621934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,2048,0.012723555167516073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3584,64,0.003920000046491623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,3072,0.024140444066789415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,2560,0.011413333316644033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,8192,0.0425235562854343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,2048,0.01035911093155543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,7168,0.03869422276814779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,1024,0.007968000239796108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,1536,0.009697777529557547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,768,0.006650666809744305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,1536,0.015416888727082146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,1024,0.008850666383902231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,512,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,1024,0.011759999725553723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,768,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,256,0.004554666578769684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,768,0.010414222048388587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,512,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,128,0.0038355555799272326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,512,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,256,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,64,0.0035582222044467926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,256,0.008651555412345463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,128,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,32,0.003527111063400904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,128,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,64,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,3072,3072,0.012774222426944308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,65536,0.23899555206298828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,32,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,3072,1536,0.010584889186753167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,2560,0.02107200026512146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,51200,0.1867013374964396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,3072,2048,0.017659554878870647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,65536,0.14244000116984049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,65536,0.3601120048099094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,51200,0.11390933725568984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,12288,0.053951111104753285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,10240,0.04770222306251526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,16384,0.03748888770739237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,51200,0.28285421265496147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,12288,0.03033155534002516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,8192,0.03671822283003066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,16384,0.0926097763909234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,10240,0.027020444472630818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,7168,0.02920444475279914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,8192,0.022042666872342426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,6144,0.02773155603143904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,8192,0.04809244473775228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,7168,0.04256977637608846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,5120,0.024157333705160353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,6144,0.018394667241308425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,6144,0.03760177890459696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,4096,0.019481778144836426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,5120,0.0163111107216941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,5120,0.03222844335767958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,4096,0.01440355512830946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,16384,0.0744017759958903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,4096,0.027144889036814373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,3072,0.015669332610236276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,3584,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,3584,0.02443555494149526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,2560,0.013456000222100152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,3072,0.01236977759334776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,3072,0.021881777379247878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,12288,0.06811200247870551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,2048,0.011563555234008364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,10240,0.05853422482808431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,2560,0.011398221883508893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,1536,0.009208888643317753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,7168,0.020492444435755413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,2048,0.01016088906261656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,1024,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,2048,0.016195555528004963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,1536,0.009657777845859528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,768,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,1024,0.009034666750166152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,1536,0.01387288918097814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,512,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,1024,0.011030221978823343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,768,0.008301333420806462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,256,0.0042168887125121225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,768,0.009710222482681274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,512,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,128,0.003547555456558863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,512,0.008400888906584846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,256,0.0075937774446275495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,64,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,32,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,256,0.008064000142945183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2560,128,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,128,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,64,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,65536,0.2031866709391276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,32,0.007689778175618913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,16384,0.05615288681454129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,65536,0.1363697714275784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,51200,0.1883528894848294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,12288,0.04411555661095513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,51200,0.11107733514573838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,65536,0.3127937846713596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,16384,0.03595822387271457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,10240,0.03263911273744371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,51200,0.25006310145060223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,12288,0.029253333806991577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,16384,0.07790044281217787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,8192,0.027688889039887324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,12288,0.05956977605819702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,7168,0.026558222042189703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,10240,0.02699911097685496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,6144,0.022309333086013794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,8192,0.02123466630776723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,10240,0.05127999848789639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,5120,0.01853155593077342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,7168,0.01956533392270406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,8192,0.04195911023351881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2560,3584,0.017650667164060805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,4096,0.015124445160230001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,7168,0.037885334756639265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,5120,0.015814221567577787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,6144,0.03355733222431607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,4096,0.013761777844693927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,5120,0.02879377868440416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,3072,0.012187555432319641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,4096,0.0239991115199195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,3584,0.012739555703269111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,2560,0.010780444575680627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,3584,0.021887999441888597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,3072,0.011732444167137146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,2048,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,3072,0.019680000013775296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,2560,0.011074666347768573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,2560,0.016920000314712524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2560,2560,0.018944889307022095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,1536,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,1024,0.005670222143332164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,2048,0.010015111002657149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,2048,0.014860444598727755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,1536,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,768,0.005128889034191768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,1024,0.00870666652917862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,1536,0.01239377756913503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,512,0.004306666553020477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,1024,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,768,0.008015111088752747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,256,0.003528000166018804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,512,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,128,0.0031973332580592898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,256,0.007345777418878343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,512,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,64,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,256,0.007333333293596904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,32,0.0029084444459941653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,128,0.0074346669846110875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,128,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,64,0.007093333535724216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,65536,0.20308532979753283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,32,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,51200,0.15669066376156277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,65536,0.09220888879564072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,65536,0.26426222589280873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,16384,0.05584088961283366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,51200,0.07288355297512479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,51200,0.20683911111619738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,12288,0.042062222957611084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,16384,0.02535555594497257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,10240,0.033490667740503945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,16384,0.06685955656899346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,2048,768,0.008962666822804345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,12288,0.023798222343126934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,8192,0.027151111099455092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,2048,6144,0.017613333132531907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,10240,0.019525332583321463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,2048,3584,0.013963555296262106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,10240,0.04396533303790622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,8192,0.01585422290696038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,6144,0.020284444093704224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,8192,0.03621600071589152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,7168,0.014342221948835584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,5120,0.017681777477264404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,6144,0.013258666627936892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,7168,0.033055110110176936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,4096,0.015096000499195524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,5120,0.012064889073371887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,6144,0.029096888171301946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,3584,0.013215111361609565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,3072,0.011730666789743634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,5120,0.024918221765094336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,4096,0.010737777584128909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,3584,0.01035377797153261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,4096,0.02113155523935954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,3072,0.010409778190983666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,2048,0.008630221916569604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,2560,0.009337777892748514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,3072,0.01703644461101956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,1536,0.007132444116804335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,2560,0.014902222487661572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,2048,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,1024,0.005711110929648082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,1536,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,2048,0.013067555924256643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,768,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,12288,0.051312890317704946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,1536,0.01108888867828581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,512,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,7168,0.023725334140989516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,1024,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,256,0.0035546666218174826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,768,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,1024,0.009120889008045197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,128,0.0032017777363459268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,512,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,768,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,256,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,512,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,32,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1536,128,0.006800000038411882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,256,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,128,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,65536,0.13257600201500788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,64,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,32,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,51200,0.10860000054041545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1536,3584,0.019123555885420907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,65536,0.08656799793243408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,16384,0.03238577644030253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,12288,0.024285333024130926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,65536,0.21165866321987578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,51200,0.06971377796596952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,16384,0.024023110667864483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,10240,0.02028888960679372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,51200,0.16682310899098715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,16384,0.05516889029079013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,8192,0.016880000631014507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,12288,0.04257155458132426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,10240,0.018048889107174344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,7168,0.01481066644191742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,10240,0.03670310974121094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,8192,0.014702222413486905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,6144,0.012868444124857584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,7168,0.01367911116944419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,64,0.0028231110837724474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,7168,0.02757599949836731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,5120,0.011000889043013254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,6144,0.012578667038016848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,4096,0.009575111170609793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,5120,0.01148000028398302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1536,2560,0.010327111515733931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,5120,0.020663999848895602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,4096,0.010348444183667501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,4096,0.01789688898457421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,3072,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,3584,0.009902222288979424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,3584,0.016181333197487723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,2560,0.006660444455014334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,3072,0.009372444616423713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,12288,0.019876443677478366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,3072,0.01479111115137736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,2048,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,2560,0.00904977818330129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,1536,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,2560,0.013112000293201871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,2048,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,8192,0.030452443493737116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,2048,0.01178311142656538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,1536,0.008037333687146505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,1024,0.007608888877762689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,1536,0.009713778065310584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,512,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,1024,0.008007110820876228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,6144,0.024375999967257183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,768,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,256,0.0028231110837724474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,768,0.007326222128338284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,512,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,128,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,512,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,64,0.002475555572244856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,128,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,256,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,32,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,128,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,65536,0.10676533646053737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,64,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,1024,32,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,51200,0.08756177955203587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,65536,0.08565955691867405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,16384,0.03180088930659824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,1024,0.004229333251714706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,51200,0.0693733361032274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,768,0.003641777982314428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,12288,0.024494222468800012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,51200,0.13556267155541316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,16384,0.023553777072164748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,10240,0.02035199933581882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,16384,0.0455022222465939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,12288,0.019668444991111755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,8192,0.017525333497259352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,10240,0.018017777138286166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,12288,0.035995556248558894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,7168,0.015545777148670621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,1024,3584,0.00850044439236323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,10240,0.030846221579445735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,6144,0.014184888866212634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,1024,256,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,5120,0.011960888902346293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,7168,0.013464888764752282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,6144,0.012499555945396423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,7168,0.02347466680738661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,4096,0.010206222534179688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,6144,0.02067466742462582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,5120,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,3584,0.0092257774538464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,5120,0.017850667238235474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,4096,0.010278222461541494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,3072,0.008346666892369589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,3584,0.009742221898502773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,4096,0.015431111057599386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,2560,0.007359111474619971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,3584,0.014041778114106921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,3072,0.009378666679064432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,2048,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,65536,0.17218311627705893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,3072,0.012660444610648684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,2560,0.009104000197516548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,2048,0.008638222184446123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,2560,0.011762667033407422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,1024,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,1536,0.008151111503442129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,1536,0.00867022242810991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,768,0.0041964443193541635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,1024,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,1024,0.007342221836249034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,512,0.0035146665241983202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,8192,0.014471999473041959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,768,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,8192,0.025785777303907607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,256,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,768,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,512,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,128,0.002536000063021978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,256,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,512,0.006254222244024277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,64,0.002507555608948072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,768,128,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,32,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,256,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,128,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,65536,0.08526311318079631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,64,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,51200,0.06990577777226765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,32,0.005628444254398346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,768,1536,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,65536,0.08568177620569865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,16384,0.025407999753952026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,65536,0.1655235555436876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,12288,0.020964443683624268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,51200,0.0692231125301785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,768,2048,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,10240,0.018523555662896898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,16384,0.02288177775012122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,51200,0.1301786634657118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,8192,0.015728889240158927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,12288,0.01925244430700938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,16384,0.044059554735819496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,10240,0.017423111531469557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,7168,0.015565334094895257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,12288,0.03474666674931844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,6144,0.013056000073750814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,8192,0.014208000567224292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,10240,0.029771556456883747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,7168,0.013167999684810638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,8192,0.024587555064095393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,6144,0.012148444851239523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,7168,0.022501332892311945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,4096,0.010394666757848527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,3584,0.007055111229419708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,6144,0.019913777709007263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,5120,0.011122666299343109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,3072,0.006555555595291986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,5120,0.01740711099571652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,4096,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,2560,0.0058266665372583605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,3584,0.009657777845859528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,3072,0.009325332939624786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,3584,0.013830222189426422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,2048,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,3072,0.012519111235936483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,1536,0.004228444563017951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,2560,0.011077333655622272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,2048,0.008451555338170793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,1024,0.0035573335157500375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,1536,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,768,0.0032008888406885993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,1536,0.008989333278603023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,1024,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,512,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,1024,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,768,0.007186666958861881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,256,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,768,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,512,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,5120,0.01149866647190518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,512,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,256,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,64,0.002161777681774563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,256,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,128,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,32,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,128,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,65536,0.04502044452561272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,4096,0.014873777826627096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,64,0.005527110977305307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,51200,0.037123554282718234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,32,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,512,2560,0.008981333010726506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,65536,0.08522133032480876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,16384,0.015122666954994202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,12288,0.011220444407727985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,65536,0.1255928940243191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,512,2048,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,51200,0.09846311145358616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,10240,0.012335999972290464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,16384,0.03433955709139506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,12288,0.018923555811246235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,8192,0.01204177737236023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,12288,0.026906667484177485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,10240,0.01755288905567593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,7168,0.010997333460383944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,10240,0.02306933369901445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,8192,0.014112000664075216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,6144,0.009664889011118147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,8192,0.019888889458444383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,512,128,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,5120,0.008440888590282863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,7168,0.013092444174819522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,7168,0.01794933279355367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,4096,0.007251555720965068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,5120,0.011095110740926532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,6144,0.016163556112183463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,3584,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,5120,0.014127999544143677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,4096,0.009983110758993361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,3072,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,3584,0.009635555247465769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,3584,0.011604444848166572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,2560,0.0053226666318045715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,3072,0.009315555294354757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,16384,0.022304000125990972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,3072,0.010601777997281818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,2560,0.008973333570692275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,2560,0.009379555781682333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,1536,0.004192000048028098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,2048,0.008349333372380998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,2048,0.008990222381220924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,1024,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,1536,0.00793422261873881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,1536,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,6144,0.012076444096035428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,768,0.0031724443866146933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,1024,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,512,0.0030382221771611106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,768,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,768,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,4096,0.012477333347002665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,256,0.0025235555238193935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,512,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,128,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,256,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,64,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,256,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,128,0.00657155571712388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,32,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,128,0.005156444592608346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,64,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,256,32,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,51200,0.027457776996824477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,65536,0.08497333526611328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,65536,0.12445155779520671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,1024,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,16384,0.012083555261294046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,51200,0.06820888651741876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,51200,0.09801066584057277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,16384,0.022058667408095464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,12288,0.010115555591053432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,10240,0.00959644466638565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,16384,0.033823112646738686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,512,0.006675555474228329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,12288,0.018684445155991446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,8192,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,256,51200,0.06867555777231853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,12288,0.026590221458011206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,7168,0.008014221986134848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,8192,0.014096000128322177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,256,2048,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,6144,0.008108444511890411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,8192,0.019296000401178997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,7168,0.013088000317414602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,5120,0.007584888901975419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,65536,0.034621334738201566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,7168,0.017795556121402316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,4096,0.006907555378145642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,6144,0.016048888365427654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,5120,0.01109333336353302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,3584,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,4096,0.009988444546858469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,5120,0.014184888866212634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,3072,0.005663110978073544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,4096,0.01242844429281023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,3584,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,2560,0.0051857779423395796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,3584,0.01145333300034205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,3072,0.009143999881214565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,2048,0.004557333472702238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,3072,0.010370666782061258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,2560,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,1536,0.0038480001191298165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,2560,0.009576888548003303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,1024,0.003244444520937072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,2048,0.008757332960764566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,1536,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,768,0.0028764444092909494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,1536,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,10240,0.01717511150572035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,1024,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,512,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,1024,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,768,0.007190222541491191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,256,0.002467555511328909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,512,0.006741333338949416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,512,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,6144,0.01204711116022534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,128,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,256,0.006779555645253923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,64,0.0021066665649414062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,256,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,128,0.006593777901596493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,128,32,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,65536,0.03115733464558919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,128,0.004996444616052839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,51200,0.026100445124838088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,64,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,16384,0.011014222270912595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,32,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,12288,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1281,128,2048,0.0083155557513237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,65536,0.12393155362870957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,10240,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,16384,0.03325600094265408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,8192,0.0076657773719893555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,7168,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,12288,0.026392888691690233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,6144,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,10240,0.023012444376945496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,5120,0.007368000017272101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,8192,0.019338667392730713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,768,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,4096,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,3584,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,6144,0.015822221835454304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,3072,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,5120,0.014179555906189812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,4096,0.012320000264379712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,2560,0.004855999930037392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,2048,0.004517333375083076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,3584,0.011344888971911537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,1536,0.0038622220357259116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,3072,0.010641777680979835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,1024,0.003148444410827425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,2560,0.009331555830107795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,768,0.0028479999552170434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,2048,0.008637333081828224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,1536,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,512,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,256,0.0021964444054497615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,1024,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,51200,0.09770222504933675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,128,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,768,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,512,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,128,10240,0.023016888234350417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,256,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,65536,0.03170133299297757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,51200,0.02478844424088796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,128,0.005006222261322869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,64,0.005179555465777715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,16384,0.00908088849650489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,12288,0.011533333195580376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,7168,0.017719111508793302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,10240,0.010572444233629437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,8192,0.008847999903890822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,65536,0.12424000104268391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,51200,0.09756710794236924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,16384,0.033384889364242554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,7168,0.008387555678685507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,12288,0.02630311085118188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,6144,0.008711111214425828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,5120,0.008221333225568136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,10240,0.02270577847957611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,4096,0.007517333659860823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,8192,0.019274666905403137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,7168,0.01773155563407474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,6144,0.0158897770775689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,3072,0.006007110906971826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,5120,0.014152889450391134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,4096,0.012422222230169507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,2048,0.004455111092991299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,3584,0.011433777709801992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,1536,0.003567111161020067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,3072,0.010425777898894416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,1024,0.0031804444475306403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,2560,0.009319110876984065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,768,0.002834666727317704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,2048,0.008660444782839881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,512,0.0025048889219760895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,1536,0.007795555724038019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,256,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,1024,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,64,32,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,768,0.005967111223273807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,64,0.00180177783800496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,512,0.005717333406209946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,256,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,128,0.005067555440796746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,64,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,3584,0.006730666591061487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1281,32,32,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,2560,0.005486222190989389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,64,32,0.0018702221827374564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1281,32,128,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,16384,1.5742870966593425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,16384,1.7879369523790147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,16384,0.5076364411248101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,51200,1.8397884368896484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,12288,1.1551893022325304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,12288,1.2564666536119249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,16384,1.520751953125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,10240,0.9416364034016927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,51200,4.940778520372179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,10240,1.0359172821044922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,12288,0.4503475295172797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,12288,1.1273635228474934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,10240,0.32956711451212567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,8192,0.8604515393575033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,8192,0.27742311689588756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,51200,6.169565412733291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,7168,0.6605626742045084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,10240,0.8885422282748752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,8192,0.8193306922912598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,7168,0.23970399962531197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,6144,0.6412106619940864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,8192,0.7123760117424859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,7168,0.7079440222846137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,6144,0.20863821771409777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,6144,0.6334213150872124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,5120,0.5150293244255914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,5120,0.1776204506556193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,4096,0.34435200691223145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,6144,0.5355688730875651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,5120,0.5477573076883951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,4096,0.1466791099972195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,3584,0.35364709960089785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,5120,0.4642000198364258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,4096,0.4222266674041748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,3584,0.3597875436147054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,3584,0.13300355275472006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,4096,0.3673680093553331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,3072,0.11816355917188857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,3584,0.31892622841729057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,7168,0.625873777601454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,3072,0.31514043278164333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,3072,0.2879689004686144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,2048,0.18249955442216662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,2560,0.10541955629984538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,2560,0.2638915644751655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,2560,0.23952444394429526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,2048,0.08891466591093276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,2048,0.21281245019700792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,1536,0.1605795489417182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,2048,0.19996266894870332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,1024,0.1087048914697435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,1536,0.07122933202319674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,1536,0.16898488998413086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,2560,0.24238665898640951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,51200,4.96577411227756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,1024,0.056686222553253174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,768,0.08428533209694757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,1024,0.12381688753763835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,1024,0.12815022468566895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,512,0.05863733424080742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,768,0.050366222858428955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,768,0.10271644592285156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,256,0.042099555333455406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,512,0.045928001403808594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,512,0.08855288558536106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,768,0.13616889052920872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,512,0.08281599813037448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,128,0.03685155510902405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,256,0.04227022329966227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,64,0.03467822074890137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,256,0.0654933320151435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,65536,128,0.040834665298461914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,32,0.03498933381504483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,65536,3072,0.295413335164388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,256,0.0798871119817098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,128,0.05085511008898417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,128,0.07538933224148221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,32,0.07480888896518283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,65536,1536,0.15943911340501574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,65536,64,0.07434933053122626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,65536,1.83934572007921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,51200,1.3829849031236436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,51200,3.777477264404297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,16384,1.489135954115126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,65536,4.908087836371528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,51200,4.725106557210286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,16384,1.57641421424018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,16384,0.39450666639539933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,12288,1.106207953559028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,12288,0.31203556060791016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,12288,0.9859724044799805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,65536,6.6831554836697045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,16384,1.1299058066474068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,10240,0.2665688991546631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,10240,0.8010906643337674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,10240,0.8828559981452094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,12288,0.821464909447564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,8192,0.21776533126831055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,8192,0.6361955536736382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,10240,0.6882746484544543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,51200,3.9050381978352866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,7168,0.5712061987982856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,8192,0.5567546420627171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,8192,0.6297138002183702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,7168,0.1887591150071886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,6144,0.5315439965989854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,7168,0.4937973552280002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,6144,0.4983644485473633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,5120,0.3849697907765706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,6144,0.4284186628129747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,5120,0.40563199255201554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,5120,0.14237156179216173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,7168,0.5621342129177517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,4096,0.3367155657874213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,65536,5.009985605875651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,5120,0.3571084340413411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,4096,0.11638489034440781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,3584,0.24981154335869682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,4096,0.33232177628411186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,6144,0.166557338502672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,3584,0.1058186690012614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,3072,0.2568942175971137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,3584,0.3001013331943088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,2560,0.21888266669379342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,3072,0.24392265743679473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,3072,0.09485422240363227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,3072,0.22729155752393934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,3584,0.3373093340131972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,2560,0.08243378003438313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,2560,0.21590488486819795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,2560,0.19266933865017363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,1536,0.13484266069200304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,2048,0.1683724456363254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,2048,0.0699288911289639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,1024,0.08346577485402425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,2048,0.15992977884080675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,1536,0.05707822243372599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,1536,0.13524266084035239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,768,0.06665688753128052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,1024,0.10248088836669922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,1024,0.04544088906712002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,512,0.05154844456248813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,1024,0.09822311004002889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,4096,0.29073333740234375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,768,0.08189511299133301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,256,0.03414577907986111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,512,0.037583112716674805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,512,0.07017422384685941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,512,0.06790755854712592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,2048,0.14684889051649305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,128,0.030107554462220933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,256,0.063646223809984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,256,0.034550223085615374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,256,0.0540231102042728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,64,0.02460711035463545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,128,0.03346488873163859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,51200,32,0.025116443634033203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,128,0.059578663773006864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,128,0.04154577851295471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,64,0.05949066744910347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,51200,32,0.05990488661660088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,1536,0.12753244241078696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,51200,768,0.04153866569201151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,51200,768,0.08316444688373141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,65536,0.5777706570095485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,51200,0.4595031208462185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,65536,1.6058230929904516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,51200,1.2590950859917533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,16384,0.44269598854912656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,16384,0.13566577434539795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,16384,0.4814604653252496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,65536,1.9808533986409504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,12288,0.10920000076293945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,12288,0.31535911560058594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,16384,0.3740728961096869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,12288,0.2811777856614855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,10240,0.28475112385219997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,10240,0.26360710461934406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,10240,0.09205422136518691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,51200,1.2380586200290256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,10240,0.23931643697950575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,51200,1.5883412890964086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,12288,0.3401510980394151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,7168,0.19816354910532633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,8192,0.08191200097401936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,65536,1.6318204667833116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,8192,0.20661066638098824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,6144,0.1734231048160129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,7168,0.06679822338951959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,5120,0.15016533268822566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,6144,0.05945777893066406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,6144,0.20610400040944418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,7168,0.17038222153981528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,6144,0.14725600348578557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,4096,0.12365333239237468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,5120,0.054140445258882314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,5120,0.13457955254448786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,8192,0.22148356172773573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,4096,0.10729155275556777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,4096,0.0428693327638838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,5120,0.1288248830371433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,4096,0.1060462196667989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,3072,0.08561066786448161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,3584,0.09511644310421413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,3584,0.03792000148031447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,8192,0.1908808946609497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,7168,0.18413333098093668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,3584,0.09338044457965428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,3072,0.033899555603663124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,2560,0.0766071081161499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,3072,0.08324178059895833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,2048,0.062436441580454506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,3072,0.08530133300357395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,2560,0.030543999539481267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,1536,0.04469244347678291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,2560,0.07219022512435913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,2560,0.07226488987604777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,2048,0.02599999970859951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,2048,0.05977421998977661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,1024,0.03282311227586534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,2048,0.06262933545642428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,1536,0.021148444877730474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,1536,0.04799555407630073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,1024,0.017815111411942374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,1536,0.051606221331490405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,3584,0.1085404422548082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,1024,0.036657777097490095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,512,0.018010665973027546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,1024,0.04189866781234741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,768,0.01633066601223416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,768,0.03024622135692173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,256,0.012863111164834766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,512,0.015080889066060385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,512,0.024666666984558105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,128,0.010084444450007545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,512,0.03048711021741231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,256,0.013783999615245394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,256,0.021736888421906367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,64,0.008626666333940294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,256,0.02424266603257921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,32,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,128,0.021271111236678228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,128,0.01869333287080129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,64,0.021306667062971327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,16384,32,0.021316443880399067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,16384,768,0.024297777149412367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,16384,768,0.03558844327926636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,65536,0.4395039876302083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,16384,128,0.013561777770519257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,51200,1.0211750666300456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,65536,1.392571555243598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,65536,1.458006223042806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,16384,0.3295973406897651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,51200,0.3453217877282037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,51200,1.1002178192138672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,12288,0.2353422111935086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,16384,0.11243289046817356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,16384,0.30097511079576283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,12288,0.08371555805206299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,12288,0.2609911229875353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,12288,0.22214310699039033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,10240,0.07180088758468628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,8192,0.1700977749294705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,10240,0.2079075574874878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,65536,1.2754667070176866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,10240,0.18556888898213705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,8192,0.05996266338560316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,7168,0.17464710606469047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,8192,0.16141600079006618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,7168,0.1406222184499105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,6144,0.14163733853234187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,8192,0.15182311005062527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,7168,0.0579893324110243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,16384,0.3570862346225315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,5120,0.12648710939619276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,6144,0.04673511121008131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,7168,0.13609866301218668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,10240,0.2236417664421929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,6144,0.12134222189585368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,4096,0.0925262239244249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,5120,0.10304622517691718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,5120,0.039956443839603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,6144,0.11999821662902832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,3584,0.08420177963044907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,4096,0.03419377737575107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,5120,0.10452177789476182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,4096,0.0834675563706292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,3072,0.06369866927464803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,3584,0.030065778228971694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,4096,0.08590577708350287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,3584,0.07399733199013604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,2560,0.05972533093558418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,3584,0.07772355609469943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,3072,0.027082666754722595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,2560,0.024728000164031982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,51200,0.9252924389309354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,3072,0.06843466891182794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,3072,0.06463555494944255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,2560,0.0569795568784078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,1536,0.03607911202642653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,2560,0.059008002281188965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,2048,0.021389333738221064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,2048,0.04671022295951843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,1024,0.02351466649108463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,1536,0.017596445149845548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,2048,0.05176888902982076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,1536,0.03785066803296407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,768,0.020598222812016804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,1536,0.04262666569815742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,1024,0.014979556202888489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,1024,0.028919110695521038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,512,0.01704266667366028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,768,0.01386044422785441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,768,0.02424088948302799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,256,0.013162666724787818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,768,0.02994222111172146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,512,0.01976444489426083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,512,0.0251528885629442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,128,0.011343999869293638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,256,0.011844444606039258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,256,0.017601778109868366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,2048,0.047091556919945605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,64,0.010615111225181155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,256,0.020607110526826646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,12288,32,0.010823999842007955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,128,0.011914666328165265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,128,0.01719199948840671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,128,0.01685422162214915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,64,0.01720355616675483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,12288,32,0.01720177796151903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,12288,1024,0.034471111165152654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,12288,512,0.012828444441159567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,65536,0.37214843432108563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,65536,0.9947502348158094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,51200,0.804456869761149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,16384,0.29399998982747394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,51200,0.30260001288519967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,51200,0.9480453067355685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,65536,1.189841800265842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,16384,0.09302488962809245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,16384,0.29068533579508465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,12288,0.2121244404051039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,10240,0.1701271136601766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,16384,0.2709626621670193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,12288,0.07340533203548856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,12288,0.2103795607884725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,12288,0.19561155637105307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,8192,0.13257955180274114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,10240,0.1828275521596273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,10240,0.06362133555942111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,7168,0.11274666256374782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,8192,0.04990577697753906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,10240,0.16366488403744167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,51200,0.8200320137871636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,6144,0.10142133633295695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,7168,0.12064354949527317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,7168,0.04469866553942362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,5120,0.08745955758624607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,7168,0.12076711654663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,6144,0.12121333016289605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,6144,0.1048275563451979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,5120,0.03627466824319627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,4096,0.06755910979376899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,5120,0.0885395540131463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,5120,0.08990133470959133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,4096,0.029410666889614526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,3584,0.059402664502461754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,4096,0.07278044356240167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,4096,0.07591911156972249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,3072,0.051078223519855075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,3584,0.026781333817376032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,65536,1.076559066772461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,3584,0.06412977642483182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,2560,0.04518933428658379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,3584,0.06743466854095459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,8192,0.13479910956488714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,3072,0.023997333314683702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,2048,0.03590755661328634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,3072,0.05605600277582804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,2560,0.02180622186925676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,3072,0.060699555608961314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,6144,0.0395066671901279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,2560,0.048932443062464394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,1536,0.02754755483733283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,2560,0.05305244525273641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,2048,0.019015999303923715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,1024,0.018632888793945312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,2048,0.04028266668319702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,1536,0.01663644446267022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,2048,0.04625155528386434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,768,0.014967999524540372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,1536,0.03300977746645609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,1536,0.03846755623817444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,512,0.012988444831636218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,1024,0.025383111503389146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,1024,0.030525333351559106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,768,0.021007999777793884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,768,0.026495110657480028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,256,0.01016355554262797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,512,0.017274666163656447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,8192,0.13725244998931885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,512,0.023232888844278123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,128,0.008623110751310984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,256,0.01106844428512785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,256,0.015634665886561077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,256,0.019500444332758587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,128,0.01107377807299296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,128,0.015132443772421943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,32,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,10240,128,0.01550222271018558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,64,0.01516888870133294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,10240,32,0.015094222293959724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,768,0.012759110993809171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,512,0.011732444167137146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,65536,0.8136204613579644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,65536,0.3249582184685601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,10240,64,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,65536,0.9667804506089953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,51200,0.6546115345425075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,10240,1024,0.014000000225173103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,16384,0.208752883805169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,51200,0.25301512082417804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,12288,0.1836400032043457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,16384,0.08505422539181179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,16384,0.25421155823601616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,12288,0.1624737713072035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,10240,0.15162221590677896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,16384,0.21703378359476724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,51200,0.669106642405192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,10240,0.05217511124081082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,10240,0.13678399721781412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,8192,0.12944532765282526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,10240,0.14735644393497044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,7168,0.10854844252268474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,8192,0.04434222314092848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,8192,0.11554843849605984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,51200,0.7486133575439453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,7168,0.11001511414845784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,7168,0.0376853346824646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,6144,0.08431910806232029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,65536,0.9130018022325305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,12288,0.062333332167731396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,5120,0.08516533507241143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,6144,0.033767110771603055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,6144,0.08752710951699151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,12288,0.18178933196597627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,6144,0.09017422464158799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,5120,0.07321511374579535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,5120,0.02966133422321743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,3584,0.05565510855780708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,4096,0.025057777762413025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,4096,0.06085244152281019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,8192,0.11164000299241807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,3072,0.05002489023738437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,4096,0.06544978088802762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,3584,0.023010666171709698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,3584,0.05376533336109585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,2560,0.04042222102483114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,3584,0.05888977977964613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,3072,0.02124711043304867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,7168,0.09954755836062962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,3072,0.04691288868586222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,2048,0.03163822160826789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,2560,0.019271110494931538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,2560,0.0411137772930993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,1536,0.022950222094853718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,2048,0.017263111140992906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,2048,0.03421688742107815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,2048,0.040214220682779946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,1024,0.017072889539930556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,4096,0.06210044357511732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,1536,0.014390221900410123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,1536,0.0281644430425432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,768,0.013419555293189155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,1536,0.03382755650414361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,1024,0.012516444755925072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,5120,0.07848622401555379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,1024,0.021746666895018682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,512,0.010397333237859938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,1024,0.02716355522473653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,768,0.011528000235557556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,768,0.018200889229774475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,256,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,512,0.010770666930410596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,768,0.023788443870014612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,128,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,512,0.020285333196322124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,256,0.010220444036854638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,256,0.017431111799346078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,3072,0.05307555529806349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,64,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,8192,128,0.010268444816271463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,8192,32,0.007522666619883643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,128,0.013122667041089801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,128,0.013914666242069669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,8192,2560,0.04654844601949056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,64,0.01310222182008955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,32,0.013110222087966071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,65536,0.7276275422838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,65536,0.26724621984693736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,51200,0.5247093306647407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,512,0.01484266585773892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,8192,256,0.013464888764752282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,16384,0.1752986643049452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,51200,0.20695466465420195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,16384,0.19327110714382598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,16384,0.0686640010939704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,16384,0.18311556180318198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,65536,0.7246355480617948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,12288,0.1366693311267429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,51200,0.6897635459899902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,10240,0.1186622248755561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,12288,0.05481777919663323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,12288,0.1479075617260403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,51200,0.5673226780361599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,10240,0.1248319943745931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,8192,0.09699822134441799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,10240,0.04706844356324938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,7168,0.08556711011462742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,10240,0.12099111080169678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,65536,0.8510906431410047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,8192,0.036159111393822566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,8192,0.10689955287509495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,6144,0.075527999136183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,7168,0.03264088763131036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,7168,0.09069510963228013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,7168,0.09018577469719781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,5120,0.06124711036682129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,6144,0.07690222395790948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,6144,0.0777626633644104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,4096,0.048680888281928174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,5120,0.02573777735233307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,5120,0.06599999798668756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,5120,0.06815289126502143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,3584,0.04186488853560554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,4096,0.021874666213989258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,4096,0.053871999184290566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,4096,0.057012445396847196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,3584,0.020220445262061227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,3072,0.032009777095582746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,3584,0.048320889472961426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,3584,0.052536888255013355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,3072,0.01854222185081906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,3072,0.04284088810284933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,8192,0.09947378105587429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,2048,0.02474400069978502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,2560,0.017257778180970084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,2560,0.0374951097700331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,6144,0.02937777837117513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,2560,0.04039644532733493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,1536,0.01925155520439148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,2048,0.01512888901763492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,2048,0.031166222360399034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,1024,0.014000000225173103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,2048,0.03544888893763224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,1536,0.013167999684810638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,1536,0.02558044426971012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,768,0.011164444188276926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,1536,0.030084444416893855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,1024,0.011362666885058085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,1024,0.01977244516213735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,512,0.008936000367005667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,12288,0.14235022332933214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,768,0.010558221903112201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,768,0.016488000750541687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,256,0.006457777900828256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,768,0.02130577796035343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,2560,0.0325857765144772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,512,0.01000711073478063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,128,0.00612088872326745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,512,0.013795555465751223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,512,0.018277333842383493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,3072,0.04605422086185879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,64,0.00535111129283905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,256,0.012464889221721225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,256,0.016367110941145156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,7168,32,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,128,0.00962755580743154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,128,0.012168000141779581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,128,0.01351466692156262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,64,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,7168,32,0.012048888537618848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,65536,0.5749955707126194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,7168,1024,0.023423110445340473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,51200,0.4284186628129747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,16384,0.15369778209262425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,65536,0.7347182167900933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,7168,256,0.009308444129096137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,51200,0.19695999887254503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,51200,0.5765013164944118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,12288,0.1287013292312622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,16384,0.06304711103439331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,16384,0.17778666814168295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,65536,0.7027289072672526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,16384,0.1817022164662679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,12288,0.12769867314232722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,10240,0.11559732755025227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,12288,0.04945422212282816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,8192,0.09464266565110947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,12288,0.15781511200798884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,10240,0.04296888907750448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,10240,0.10839555660883586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,51200,0.5571280055575901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,65536,0.2520986662970649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,7168,0.08162311050626966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,8192,0.03456177645259433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,6144,0.06503822406133015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,7168,0.0314355558819241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,7168,0.077711111969418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,8192,0.10067644384172227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,7168,0.08652622169918484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,5120,0.05794488721423679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,6144,0.02792088852988349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,6144,0.06758399804433186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,6144,0.0768231087260776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,5120,0.025082666012975905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,5120,0.05813066826926338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,5120,0.06592977709240384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,3584,0.041690667470296226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,4096,0.0215022216240565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,4096,0.047468443711598717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,4096,0.05568089087804159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,3072,0.03292355603641934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,3584,0.019855111837387085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,3584,0.04176533222198486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,3584,0.050253334972593516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,10240,0.11796089013417561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,2560,0.02947288751602173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,3072,0.01812088820669386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,8192,0.08789955245123969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,3072,0.0369484457704756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,3072,0.04524088899294535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,2560,0.0166302224000295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,2560,0.0326257778538598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,1536,0.0196088883611891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,2560,0.039742221434911094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,2048,0.014832889040311178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,2048,0.027824888626734417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,1024,0.014460444450378418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,4096,0.042265776130888194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,2048,0.034408890538745456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,1536,0.012950222525331708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,1536,0.022636445032225713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,768,0.011904888682895236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,1536,0.02885066800647312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,1024,0.011337777806652917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,512,0.009135999613338048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,1024,0.01720711092154185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,768,0.010393777655230628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,1024,0.02332177758216858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,256,0.007183999651008182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,768,0.015089778436554803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,512,0.009710222482681274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,512,0.012447111308574677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,256,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,512,0.01789777808719211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,256,0.011422221859296164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,64,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,256,0.015116444892353482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,6144,128,0.009119999905427298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,32,0.005836444596449534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,128,0.011042666931947073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,2048,0.024315555890401203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,128,0.012866666747464074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,64,0.011081777513027191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,6144,32,0.011049778097205691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,65536,0.4665546417236328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,65536,0.22991644011603463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,51200,0.4167742199367947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,16384,0.14162755012512207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,65536,0.6170568996005589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,6144,768,0.02029600077205234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,51200,0.18208800421820748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,51200,0.4858791033426921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,12288,0.10421155558692084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,16384,0.057612445619371205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,65536,0.5674239794413248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,12288,0.11104533407423232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,12288,0.0459555553065406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,10240,0.09172355466418797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,51200,0.4489813380771213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,12288,0.11394755045572917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,10240,0.044554667340384595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,10240,0.09301244550281101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,10240,0.09819377793206109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,8192,0.07553866836759779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,8192,0.0324942237801022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,8192,0.08503822485605876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,6144,128,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,7168,0.029366221692827012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,7168,0.06717244121763441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,6144,0.05573155482610067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,7168,0.07205955849753486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,5120,0.045408000548680626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,6144,0.025295999315049913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,6144,0.05847466654247708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,16384,0.14843111568027073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,6144,0.06347733073764376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,4096,0.036200887627071805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,5120,0.04939555459552341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,5120,0.021886222892337378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,8192,0.07067644596099854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,3584,0.02886755598915948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,5120,0.05464355481995476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,4096,0.01959377858373854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,4096,0.041366222831938006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,3072,0.027199111051029626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,4096,0.04704533351792229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,7168,0.06365244256125556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,3584,0.03649866580963135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,2560,0.021987555755509272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,3584,0.04217155443297493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,3072,0.01627466744846768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,3072,0.03305777907371521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,2048,0.018953777021831937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,3072,0.03806133402718438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,2560,0.01459822224246131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,2560,0.028913779391182795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,1536,0.014774221513006421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,16384,0.16694222556220162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,2048,0.013372444444232516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,2048,0.02504355543189579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,1024,0.011041777829329172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,2048,0.029658665259679157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,1536,0.020032889313167997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,1536,0.025358221597141687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,1024,0.015445333388116626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,768,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,1024,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,512,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,1024,0.020291555258962844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,768,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,3584,0.017972444494565327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,768,0.013396444420019785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,256,0.005534222142563925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,768,0.018106665876176622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,512,0.008966222405433655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,512,0.011360888679822287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,128,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,512,0.017270222306251526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,256,0.008904889225959778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,64,0.00461777796347936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,256,0.010413332945770687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,5120,32,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,128,0.008742222355471717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,128,0.010050666828950247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,128,0.011711111499203576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,2560,0.03363644414477878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,64,0.010060444474220276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,5120,32,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,5120,1536,0.011383111278216044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,65536,0.18725511762830946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,65536,0.5112249056498209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,51200,0.4002888997395833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,16384,0.14532355467478433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,51200,0.14765955342186823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,16384,0.139573335647583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,16384,0.04900088906288147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,51200,0.43317333857218426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,12288,0.09737777709960938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,16384,0.17389599482218424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,10240,0.0764204462369283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,12288,0.03945066531499227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,12288,0.09258133172988892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,12288,0.11293421851264106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,10240,0.03403822249836392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,8192,0.05910488631990221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,10240,0.09567911095089382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,65536,0.5024302270677354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,8192,0.02789155642191569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,8192,0.06411111354827881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,7168,0.05019911130269369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,8192,0.07954133219189115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,51200,0.390334235297309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,6144,0.04368355539109972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,7168,0.02514933380815718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,7168,0.05724000268512302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,7168,0.06999911202324761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,6144,0.022639110684394836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,6144,0.049598223633236356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,5120,256,0.014455111490355598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,4096,0.030390222867329914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,6144,0.06287022431691487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,5120,0.020326221982638042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,5120,0.042584889464908175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,3584,0.024897777371936377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,4096,0.017530666457282174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,4096,0.035360889302359685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,10240,0.08078755272759332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,4096,0.04561600089073181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,3072,0.023966223001480103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,3584,0.01629155543115404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,3584,0.031650665733549334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,2560,0.020057777563730877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,3584,0.04162044326464335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,3072,0.028530667225519817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,3072,0.03715288970205519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,2560,0.024987555212444727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,2560,0.013564444250530668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,2048,0.016407110624843173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,5120,0.038570665650897555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,2560,0.03296800123320685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,1536,0.012077333198653327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,2048,0.012420444852775998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,2048,0.02109066645304362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,2048,0.02873777680926853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,1024,0.009116444322797988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,1536,0.017218665944205392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,5120,0.0540284448199802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,1536,0.024646222591400146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,1024,0.009639999932712978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,768,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,1024,0.013442666994200813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,1024,0.01945688823858897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,512,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,3072,0.014819555812411837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,768,0.011863999896579318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,768,0.01758666667673323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,256,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,512,0.00868088917599784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,512,0.010210666391584609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,65536,0.5521697998046875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,128,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,512,0.015511110424995422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,256,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,256,0.009565333525339762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,64,0.004507555729813046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,4096,32,0.00453599997692638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,256,0.013399110900031196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,128,0.0090506664580769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,4096,128,0.011232000258233813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,64,0.009004444711738164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,1536,0.011157333023018308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,4096,32,0.009010666774378883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,65536,0.3316560056474474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,65536,0.14900622102949354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,768,0.009154666629102495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,65536,0.45319202211168075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,16384,0.08771733442942302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,51200,0.11877155303955078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,51200,0.35884621408250594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,12288,0.06832889053556654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,65536,0.4199715455373128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,16384,0.03892622060245938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,16384,0.11011555459764268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,51200,0.33725865681966144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,16384,0.1127119991514418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,10240,0.05036977926890055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,12288,0.03095910946528117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,4096,128,0.00832177781396442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,12288,0.08423378070195515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,8192,0.05097155438529121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,10240,0.027469333675172593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,10240,0.07236444287829928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,8192,0.022276444567574397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,7168,0.043770667579438954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,10240,0.0746871100531684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,8192,0.05832266807556152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,6144,0.03420622150103251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,7168,0.0203155560625924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,8192,0.06361599763234456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,51200,0.3012675444285075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,5120,0.02972444560792711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,7168,0.051631109582053296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,6144,0.018716444571812946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,7168,0.05641244517432319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,4096,0.02553422252337138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,5120,0.016753777861595154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,6144,0.04958044489224752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,5120,0.03908266623814901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,3584,0.022218666142887537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,5120,0.04394488864474826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,4096,0.014491554763582019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,4096,0.032904889848497175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,3072,0.01885066595342424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,4096,0.037434667348861694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,3584,0.01367911116944419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,3584,0.029395557112163965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,2560,0.01685422162214915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,3072,0.012721777790122561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,3072,0.026551110876931086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,2048,0.014416888356208801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,3072,0.030809776650534734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,12288,0.09483200311660767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,2560,0.011366222467687396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,2560,0.023057777020666335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,1536,0.011750222080283694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,2560,0.027464000715149775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,2048,0.01034755590889189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,2048,0.0196435550848643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,1536,0.00958577791849772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,2048,0.024177778098318312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,1536,0.0159599996275372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,768,0.007461333440409766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,1536,0.020239111449983384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,6144,0.04579377836651272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,1024,0.01274400038851632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,1024,0.016887111796273124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,768,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,768,0.011223110887739392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,512,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,768,0.015485333071814643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,512,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,256,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,512,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,3584,0.03354133168856303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,512,0.01443822185198466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,128,0.003947555604908201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,256,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,256,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,64,0.0035546666218174826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,256,0.01241422196229299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,32,0.0038746665749284956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,128,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,128,0.00867022242810991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3584,1024,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3584,128,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,64,0.008384889198674096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3584,32,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3584,1024,0.008667555948098501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,65536,0.14338310559590658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,51200,0.2349155479007297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,16384,0.08324711190329658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,65536,0.409076452255249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,51200,0.11332621839311387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,51200,0.31368621190388996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,65536,0.4113271236419678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,12288,0.06859733661015828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,16384,0.03793511125776503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,16384,0.09942933585908677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,16384,0.11154845025804307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,10240,0.057583113511403404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,12288,0.030413332912656996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,12288,0.0760071078936259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,12288,0.08736355437172784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,10240,0.027095999982621934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,8192,0.04195644458134969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,10240,0.06432177623112996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,10240,0.07471911112467448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,7168,0.04005422194798788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,8192,0.022133333815468684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,65536,0.37137601110670304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,8192,0.05273511012395223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,6144,0.03532177872127957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,7168,0.020210666788948905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,5120,0.028463111983405218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,6144,0.018664888209766813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,7168,0.055833776791890465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,6144,0.04153155618243747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,4096,0.02461155586772495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,6144,0.049792889091703624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,5120,0.016427555018001132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,5120,0.035291555855009295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,3584,0.021576889687114294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,5120,0.04279822111129761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,4096,0.014479999740918478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,51200,0.3246764341990153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,4096,0.030187557140986126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,3072,0.01921422282854716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,4096,0.036658667855792575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,3584,0.01312355531586541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,2560,0.016866667403115165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,3584,0.026995556222067937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,3072,0.012623111406962076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,3584,0.03334844443533156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,3072,0.023927999867333308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,2048,0.014223999447292753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,3072,0.030351112286249798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,2560,0.011438222395049201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,2560,0.020563556088341605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,1536,0.012177777787049612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,2560,0.026995556222067937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,2048,0.017799999978807237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,1024,0.00869777798652649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,7168,0.04686311218473646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,2048,0.0240666667620341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,1536,0.009415110780133141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,1536,0.01482488877243466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,768,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,1536,0.01963644391960568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,1024,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,512,0.006153777655627992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,1024,0.011759111450778114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,1024,0.01683999929163191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,768,0.010363555616802638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,256,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,768,0.015501333607567681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,512,0.00905955582857132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,512,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,128,0.0042088888585567474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,512,0.014098667436175875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,256,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,256,0.008369777765538957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,64,0.0038417776425679526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,256,0.012415111064910889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,128,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,3072,32,0.004599110947714912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,128,0.010067555639478896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,2048,0.01035822182893753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,64,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,3072,8192,0.06192444430457222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,32,0.008013333711359236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,65536,0.2865271038479275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,65536,0.1383297840754191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,51200,0.23963022232055664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,3072,768,0.00832799987660514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,65536,0.3540648884243435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,16384,0.0707813302675883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,51200,0.10996888743506537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,51200,0.2967262268066406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,65536,0.40791111522250706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,12288,0.05595466825697157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,16384,0.08664621909459432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,51200,0.32132445441352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,12288,0.029703110456466675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,12288,0.06942133108774821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,12288,0.0874755581219991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,10240,0.026644445127911035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,10240,0.056742224428388804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,8192,0.03726666503482395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,10240,0.07394755548901029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,8192,0.021431111627154883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,8192,0.04690222276581658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,7168,0.03212533394495646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,8192,0.061866667535569936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,7168,0.019732443822754752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,7168,0.04173333446184794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,16384,0.038175112671322294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,6144,0.02772355576356252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,7168,0.05527911252445645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,3072,128,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,6144,0.017846221725145977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,6144,0.037019554111692644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,4096,0.018759111563364666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,6144,0.04860622353023953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,5120,0.03202222122086419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,5120,0.04204888807402717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,3584,0.017200888858901132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,4096,0.013889777991506787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,4096,0.027062222361564636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,4096,0.036256889502207436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,3072,0.015188443991872998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,3584,0.013027555412716336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,3584,0.0247688889503479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,3584,0.03297777639495002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,2560,0.013216888739003075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,3072,0.012129777835475074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,3072,0.02181244393189748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,3072,0.029951999584833782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,2048,0.011187555889288584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,16384,0.11161866452958848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,10240,0.04683199856016371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,2560,0.025791999366548326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,1536,0.009115555220180089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,2048,0.016348444753222995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,2048,0.010246222217877706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,2048,0.02297244469324748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,1024,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,1536,0.013778666655222574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,1536,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,1536,0.019495111372735765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,1024,0.008616888688670265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,768,0.005966222120655908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,1024,0.011006222003036074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,1024,0.016557332542207506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,512,0.004969777746333016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,768,0.008341333104504479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,768,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,768,0.015218666858143277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,256,0.003891555385457145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,512,0.00794488853878445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,512,0.008424888882372115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,2560,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,2560,0.018568888306617737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,128,0.003536000019974179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,512,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,256,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,256,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,128,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,256,0.012536889149083031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2560,5120,0.01608000033431583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,32,0.003216000066863166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,128,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2560,128,0.010044444766309526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,64,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,65536,0.23126043213738334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2560,32,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,65536,0.13240710894266763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,51200,0.17761154969533285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,65536,0.30479823218451607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,16384,0.05866310993830363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,51200,0.10502222511503433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,51200,0.23718044492933485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,65536,0.4092115561167399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,12288,0.044101334280437894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,16384,0.035242666800816856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,16384,0.07655821906195746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,51200,0.32109154595269096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,5120,0.02278577784697215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,10240,0.03763022356563144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,12288,0.028507557180192735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2560,64,0.0031973332580592898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,8192,0.029707554313871596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,10240,0.025598223010698955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,10240,0.05056888858477274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,10240,0.0725146664513482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,7168,0.026351999905374315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,8192,0.020818667279349435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,8192,0.041646222273508705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,8192,0.06119910875956217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,7168,0.018958222534921434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,6144,0.022522666388087805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,7168,0.037568890386157565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,7168,0.05400266581111484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,5120,0.019696000549528334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,6144,0.017264889346228707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,6144,0.03300266795688205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,6144,0.047912001609802246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,4096,0.01648888819747501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,5120,0.015174221661355762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,5120,0.028660443094041612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,5120,0.041843556695514254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,12288,0.058619558811187744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,3584,0.014671999547216626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,4096,0.01348533315791024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,4096,0.023898666103680927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,12288,0.08527288834253947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,3072,0.013216888739003075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,4096,0.03554933269818624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,3584,0.012561777399645912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,2560,0.011983111500740051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,3584,0.021253334151373968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,3584,0.03177511029773288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,3072,0.011736888852384357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,3072,0.019343111250135634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,2048,0.009905777871608734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,3072,0.02889510989189148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,2560,0.010942222343550788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,2560,0.01667288939158122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,1536,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,2560,0.025394666526052687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,2048,0.009902222288979424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,1024,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,2048,0.014927999840842353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,1536,0.009328000247478485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,2048,0.02254399988386366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,768,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,1536,0.01238044434123569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,1024,0.008431110945012834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,1536,0.01910044418440925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,512,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,16384,0.11292533079783122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,768,0.008001777860853408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,1024,0.016182222300105624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,256,0.003552888830502828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,768,0.008991999758614434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,768,0.014832889040311178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,128,0.0032017777363459268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,512,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,256,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,512,0.013801777528391944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,256,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,256,0.012106666962305704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,128,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,32,0.0038364442686239877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,128,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,2048,128,0.009716444545321995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,64,0.0069439998931354946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,32,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,51200,0.14822310871548122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,65536,0.2569173441992866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,65536,0.26846133338080513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,2048,1024,0.010045333041085137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,16384,0.04612000121010674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,51200,0.20304889149136016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,2048,512,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,51200,0.06952444712320964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,12288,0.03563111027081808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,51200,0.2121573289235433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,16384,0.02479377720091078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,16384,0.06569688849978976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,10240,0.029113776153988306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,12288,0.02039377722475264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,12288,0.05095377895567152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,65536,0.15857422351837158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,12288,0.05911733044518364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,8192,0.023765333824687537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,10240,0.04347822070121765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,10240,0.01853155593077342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,7168,0.02197333342499203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,65536,0.08738222387101914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,8192,0.014869333969222175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,8192,0.03609155615170797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,6144,0.01863822175396813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,8192,0.043868445687823825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,7168,0.013817778064144982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,7168,0.03279733326699998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,7168,0.03877066572507223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,6144,0.012729778057999082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,5120,0.015962666935390897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,6144,0.02847555610868666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,6144,0.03484266665246751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,2048,64,0.003552888830502828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,5120,0.011716444459226398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,4096,0.013796444568369122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,5120,0.02476088868247138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,3584,0.012468444804350534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,5120,0.030233777231640283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,16384,0.07552178038491143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,4096,0.010621333287821876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,3072,0.011042666931947073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,4096,0.020769778225156996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,3584,0.009870222045315636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,4096,0.026341333985328674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,2560,0.009639999932712978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,3584,0.01880088945229848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,3584,0.02346311178472307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,10240,0.051197333468331226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,3072,0.016898666818936665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,2048,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,3072,0.021573333276642695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,2560,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,2560,0.014834667245546976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,1536,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,2560,0.019542222221692402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,2048,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,1024,0.00519466648499171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,2048,0.013136000268989138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,1536,0.008043555749787224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,2048,0.017664889494578045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,1536,0.011036444041464063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,768,0.004519111166397731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,1536,0.0151235560576121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,1024,0.007644444704055786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,1024,0.009135111338562435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,512,0.0038977778620190094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,1024,0.013454222016864352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,768,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,768,0.008252444366614023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,256,0.0033048888047536216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,768,0.012401777837011548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,512,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,512,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,512,0.011746666497654386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,256,0.006695999867386288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,256,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,64,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,256,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,3072,0.009411555197503831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,32,0.002842666581273079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,128,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1536,128,0.00871377769443724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,64,0.0063680000603199005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1536,32,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,51200,0.09625333547592163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,65536,0.08507466978496975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,51200,0.06873599688212077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,65536,0.2635342280069987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,16384,0.0364951127105289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,51200,0.20572977595859102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,16384,0.05505955550405714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,12288,0.02648266653219859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,16384,0.02346222268210517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,16384,0.07435822486877441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,65536,0.1342746681637234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,12288,0.019561777512232464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,10240,0.022766222556432087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,12288,0.042633778519100614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,12288,0.05784177780151367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,8192,0.018068444397714403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,10240,0.017616889543003507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,65536,0.21222310596042207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,10240,0.03635911146799723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,7168,0.016470222009552848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,10240,0.050198223855760365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,8192,0.014408888088332282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,51200,0.16594578160179987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,8192,0.03051466743151347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,6144,0.014524444937705994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,8192,0.04223555657598707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,7168,0.013377777404255338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,7168,0.027335110637876723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,5120,0.01275466630856196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1536,128,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,7168,0.037772443559434675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,6144,0.012266666524940066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,6144,0.023924445112546284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,4096,0.010479999913109673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,6144,0.03370489014519586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,5120,0.011391110718250275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,3584,0.009730666875839233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,5120,0.020579554968410067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1536,128,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,4096,0.010099555883142684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,3072,0.008569777839713627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,5120,0.028904000918070476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,4096,0.017549332645204332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,3584,0.009780444204807281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,4096,0.0251875552866194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,2560,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,3584,0.016081778539551627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,3072,0.009335111412737105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,3584,0.022778666681713525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,3072,0.021273778544531927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,2560,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,2560,0.012738666600651212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,1536,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,2048,0.008443555898136562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,2048,0.011393778026103973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,2048,0.017076444294717576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,1536,0.008028444316652086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,1536,0.009704888694816166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,1536,0.015223110715548197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,1024,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,768,0.003946666502290302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,1024,0.013067555924256643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,768,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,2048,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,768,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,512,0.0034924443397257063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,768,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,2560,0.01888177792231242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,256,0.0028924445311228433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,512,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,512,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,512,0.011335111326641507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,128,0.002687999978661537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,1024,0.004553777890072929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,256,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,256,0.005969777703285217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,64,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,256,0.01036800030204985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,128,0.006637333167923822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,1024,32,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,128,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,1024,128,0.008371555142932469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,65536,0.1053653359413147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,64,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,32,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,51200,0.0839573343594869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,65536,0.08458222283257379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,65536,0.1728675497902764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,16384,0.030511998467975195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,51200,0.06822399960623847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,65536,0.25288621584574383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,51200,0.1364426612854004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,12288,0.023260444402694702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,16384,0.04548799991607666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,16384,0.022900443938043382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,51200,0.19916089375813803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,10240,0.01997422178586324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,16384,0.07113689184188843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,12288,0.019374221563339233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,8192,0.016685333516862657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,12288,0.03560622202025519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,10240,0.01757333344883389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,12288,0.0557875567012363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,10240,0.03087200058831109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,7168,0.014951111541854011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,10240,0.04785955614513821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,8192,0.014102222190962898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,8192,0.025847110483381484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,1024,1024,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,6144,0.013101333545313941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,8192,0.04060711132155524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,7168,0.013242666920026144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,5120,0.0116462219092581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,7168,0.023150222169028387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,7168,0.03616088959905837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,4096,0.009573332965373993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,6144,0.02056799994574653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,5120,0.011333333121405708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,1024,3072,0.014532445205582513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,6144,0.03169599837727017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,3584,0.009144888983832465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,4096,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,4096,0.01576266686121623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,3072,0.007867555651399825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,4096,0.024319999747806128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,3584,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,3584,0.014006222287813822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,2560,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,3584,0.022035555707083807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,3072,0.009272000028027428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,3072,0.02030666669209798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,2048,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,2560,0.011390222443474663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,2560,0.018556444181336295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,2048,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,2048,0.008404444489214156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,1536,0.004850666556093427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,2048,0.01686844395266639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,1536,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,5120,0.017998221847746108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,1024,0.0042168887125121225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,1536,0.008920888933870528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,5120,0.02775555517938402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,1536,0.014491554763582019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,768,0.0038355555799272326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,1024,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,1024,0.007351111206743453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,1024,0.01276177747382058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,768,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,512,0.0031928889867332247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,768,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,768,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,256,0.0028533333291610083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,512,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,512,0.006221333311663733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,512,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,128,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,6144,0.01218311074707243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,256,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,256,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,64,0.002500444443689452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,128,0.00664533343580034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,768,32,0.002542222125662698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,768,128,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,64,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,32,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,3072,0.012787555654843649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,2560,0.008978666530715095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,65536,0.08477333519193862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,51200,0.058719999260372586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,65536,0.24897334310743544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,768,256,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,16384,0.023499555057949487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,51200,0.13080978393554688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,51200,0.06817599799897936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,768,128,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,51200,0.1973102225197686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,12288,0.01959111127588484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,16384,0.022683555881182354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,10240,0.01761866609255473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,16384,0.044108443790011935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,65536,0.06584711207283868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,12288,0.018925334016482036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,8192,0.013839110732078552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,12288,0.034430222378836736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,12288,0.054604444238874644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,10240,0.017275555266274344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,10240,0.02990666694111294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,7168,0.012423111332787408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,10240,0.047008888589011304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,8192,0.0140337778462304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,8192,0.024412444896168176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,6144,0.010999999940395355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,8192,0.03878577881389194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,7168,0.01293955577744378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,7168,0.02222577730814616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,5120,0.009741333623727163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,7168,0.034879999028311834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,6144,0.012072888513406118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,6144,0.01982933282852173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,4096,0.0086666668454806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,6144,0.031609776947233416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,5120,0.011019555230935415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,5120,0.017210667332013447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,3584,0.00720088928937912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,5120,0.027461333407296076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,4096,0.010013333625263637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,4096,0.014828445182906257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,3072,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,4096,0.023917333947287664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,3584,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,3584,0.013668444421556262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,16384,0.07008533345328437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,3072,0.009312888814343346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,3072,0.012431999875439538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,2048,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,3072,0.020247111717859905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,2560,0.008661333057615492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,2560,0.018050667312410142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,65536,0.16544532775878906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,2048,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,1536,0.004406222038798862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,2048,0.010065777434243096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,1024,0.003536888708670934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,2048,0.016301333904266357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,1536,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,1536,0.008771555291281806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,768,0.003185777821474605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,1536,0.014476444986131458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,1024,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,1024,0.012463111016485425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,512,0.0028773333049482773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,768,0.007015110717879401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,768,0.00664533343580034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,768,0.011800888511869641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,256,0.0024862223201327855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,3584,0.02199288871553209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,512,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,512,0.006224888894293044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,128,0.002415999977125062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,512,0.011392888923486074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,256,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,2560,0.01108000013563368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,256,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,64,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,256,0.01034755590889189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,128,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,128,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,32,0.002424888933698336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,65536,0.04385777645640903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,512,128,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,64,0.005422222117582957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,512,32,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,512,1024,0.00739822205570009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,65536,0.08414222134484185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,51200,0.06757511032952203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,65536,0.2469431029425727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,51200,0.09832977586322361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,16384,0.013873777455753751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,512,2560,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,12288,0.013439111411571503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,51200,0.1962062252892388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,16384,0.021807110971874658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,16384,0.034225778447257146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,10240,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,16384,0.06913422213660346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,12288,0.026891556051042344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,12288,0.05391377872890896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,8192,0.01106844428512785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,10240,0.023031999667485554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,10240,0.04524711105558607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,7168,0.010289777484205034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,8192,0.013833777772055732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,65536,0.12537777423858643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,8192,0.01977066695690155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,6144,0.010713777608341642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,8192,0.038281778494517006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,7168,0.018076444665590923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,7168,0.034303112162484065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,6144,0.011944000091817645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,6144,0.01625777781009674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,6144,0.03053599927160475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,5120,0.011027555498811932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,5120,0.014491554763582019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,4096,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,5120,0.027219555444187585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,10240,0.0170168893204795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,4096,0.009922666682137383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,3584,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,4096,0.012450666891203986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,4096,0.023362666368484497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,51200,0.03580977850490146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,3072,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,3584,0.009487110707494948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,3584,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,2560,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,3584,0.021351999706692163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,3072,0.00909511082702213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,7168,0.01273155543539259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,3072,0.010670222342014313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,2048,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,3072,0.019592000378502738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,2560,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,1536,0.0039013334446483185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,2560,0.009571555587980483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,2560,0.017853332890404593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,2048,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,2048,0.008933333059151968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,1024,0.003265777809752358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,2048,0.016407110624843173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,1536,0.007872000336647034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,1536,0.007840889195601145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,768,0.0031591111587153543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,1536,0.014108444253603617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,1024,0.007268444531493717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,512,0.00282222218811512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,1024,0.01237155579858356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,768,0.00701333334048589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,768,0.005967999911970562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,256,0.0024906665914588505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,768,0.011445333560307821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,512,0.006607999818192587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,512,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,128,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,512,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,256,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,256,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,64,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,256,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,128,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,32,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,128,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,65536,0.03462400039037069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,256,5120,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,64,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,51200,0.02796088986926609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,32,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,65536,0.08380888568030463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,65536,0.12462488810221355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,256,12288,0.018616888258192275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,16384,0.012025777664449481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,65536,0.24674312273661295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,256,1024,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,51200,0.06744355625576444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,51200,0.09791200028525458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,12288,0.009933333430025313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,16384,0.033470223347345986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,51200,0.19579733742607963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,16384,0.02165511084927453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,10240,0.009126221968068017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,12288,0.018561777141359117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,12288,0.02645688917901781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,16384,0.06872000296910603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,8192,0.008025777836640676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,12288,0.05275111065970527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,10240,0.023192889160580103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,7168,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,10240,0.04519644379615784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,8192,0.013805333111021252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,8192,0.01926488843229082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,256,128,0.007627555893527136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,6144,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,8192,0.03804355528619554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,7168,0.01277955538696713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,7168,0.017879999346203275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,5120,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,7168,0.033977776765823364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,6144,0.011759111450778114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,4096,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,6144,0.01590044465329912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,6144,0.030280888080596924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,5120,0.011063999599880643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,5120,0.014462222655614218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,3584,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,5120,0.027104000250498455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,4096,0.009979555176364051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,3072,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,4096,0.02320266597800785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,3584,0.009468444519572789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,3584,0.011529777612951068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,2560,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,3584,0.021241777473025854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,3072,0.00905511114332411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,3072,0.010397333237859938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,2048,0.004521777646409141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,10240,0.016919111212094624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,2560,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,1536,0.003839111162556542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,2560,0.009711110757456886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,2048,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,2560,0.017877333694034152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,2048,0.008621333373917473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,1536,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,1536,0.00794222205877304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,1536,0.014011555247836642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,1024,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,768,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,1024,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,1024,0.012141333685980903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,512,0.002521777732504739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,768,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,768,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,768,0.011482666763994427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,256,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,512,0.005771555420425203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,512,0.011003555523024665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,256,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,256,0.006573333508438534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,128,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,256,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,1024,0.0031928889867332247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,64,0.0020391111158662373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,2048,0.016360888878504436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,128,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,128,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,128,32,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,128,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,65536,0.03321955601374308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,51200,0.029140445921156142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,16384,0.011398221883508893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,64,0.004991111242108875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,32,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,65536,0.12482221921284993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,51200,0.09787022405200535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,10240,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,128,4096,0.012379555238617791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,16384,0.0332622230052948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,8192,0.007680888805124495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,7168,0.007225777539942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,12288,0.0262071109480328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,6144,0.006683555742104848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1280,128,512,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,10240,0.022740445203251306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,8192,0.0192239996459749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,7168,0.017609778377744887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,4096,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,3584,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,6144,0.015853333804342482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1280,128,3072,0.019732443822754752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,3072,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,5120,0.014069333672523499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,2560,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,2048,0.004250666747490565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,4096,0.012273777690198688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,3584,0.011427555647161273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,1536,0.003842666745185852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,3072,0.010423111418883005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,1024,0.003149333306484752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,768,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,2048,0.008456000023418004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,512,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,1536,0.007739555504586961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,256,0.0021564445147911706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,12288,0.009396444592210982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,1024,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,128,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,768,0.005979555762476391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,64,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,32,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,512,0.005724444571468566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,65536,0.0336168905099233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,256,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,128,0.004951999833186467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,64,5120,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,64,0.005164444446563721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,32,0.004959110998445087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,16384,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,12288,0.011770666473441653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,10240,0.01034755590889189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,65536,0.12388088968065049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,8192,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,51200,0.09741866588592529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,16384,0.0332568883895874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,7168,0.008451555338170793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,12288,0.02621511121590932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,6144,0.00777422222826216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,10240,0.022779555784331426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,5120,0.0070017774899800616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,4096,0.006041777630647023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,64,2560,0.009462222456932068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,8192,0.019259555472267997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,7168,0.017581333716710407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,3072,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,5120,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,2560,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,4096,0.01237422227859497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,2048,0.004270222038030624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,3584,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,1536,0.0035706667436493766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,3072,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,1024,0.0029928889125585556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,2560,0.009722666607962714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,768,0.003027555636233754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,2048,0.00867733359336853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,51200,0.0258942229880227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,512,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,256,0.0021537778278191886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,1536,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,128,0.0018639999131361644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,1024,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,64,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,768,0.006182222316662471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,512,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,256,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,128,0.005101333475775189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,64,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,32,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,3584,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1280,32,6144,0.015879111157523263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1280,32,32,0.0018195555441909367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,16384,0.4957715670267741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,12288,0.9354355070326063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,16384,1.3264133665296767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,16384,1.5009021759033203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,12288,0.3736906581454807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,10240,0.7950746748182508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,12288,1.0785822338528104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,10240,0.3098533418443468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,10240,0.8865804672241211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,8192,0.7141564687093099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,51200,1.7176141738891602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,8192,0.25842489136589897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,7168,0.56513335969713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,8192,0.7219342125786675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,7168,0.23873776859707305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,7168,0.5983475579155816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,6144,0.6221031083001031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,6144,0.20280799600813124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,6144,0.5219119919670953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,5120,0.37309956550598145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,51200,4.022401597764757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,4096,0.34487377272711855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,5120,0.43369777997334796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,4096,0.1408826642566257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,4096,0.36097510655721027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,3584,0.13907378249698216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,3072,0.2603351010216607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,3072,0.10782133208380805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,2560,0.2287528779771593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,3584,0.3111226558685303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,2560,0.0950160026550293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,2048,0.1806133323245578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,2560,0.22015644444359672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,5120,0.17512533399793836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,2048,0.0829555524720086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,51200,5.1484701368543835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,3584,0.27193954255845815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,1024,0.08687377638287014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,1536,0.06462755468156603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,3072,0.22585511207580566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,1536,0.14254221651289198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,768,0.07436888747745089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,1024,0.050957332054773964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,512,0.05679821968078613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,1024,0.10835999912685818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,768,0.0465280016263326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,256,0.03730310996373495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,512,0.04343466626273262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,768,0.08686666356192695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,128,0.029843555556403265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,512,0.0756817791197035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,256,0.03962577713860406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,64,0.02555466691652934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,256,0.06863466898600261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,32,0.02624977793958452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,128,0.06396710872650146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,64,0.0637306637234158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,32,0.06385422415203519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,65536,2048,0.1818693346447415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,65536,128,0.03894311189651489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,65536,1536,0.1253919998804728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,16384,0.9892782635158963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,16384,1.1771199968126085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,65536,1.7167484495374892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,16384,0.38164356019761825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,51200,1.365145789252387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,51200,3.1905494266086154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,12288,0.29182312223646373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,12288,0.6790675587124295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,10240,0.7310221989949545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,65536,4.015305836995442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,8192,0.47259998321533203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,10240,0.2607671154869927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,12288,0.909330685933431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,51200,3.9713829888237844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,8192,0.1999653312895033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,10240,0.6800595389472114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,7168,0.18050844139522979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,6144,0.3765466743045383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,7168,0.4866906801859538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,7168,0.4156844351026747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,8192,0.7006008889940051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,6144,0.16152089171939427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,5120,0.35780800713433164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,65536,5.052136739095052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,4096,0.2871849007076687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,5120,0.13367377387152776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,6144,0.4133297867245144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,5120,0.37583645184834796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,4096,0.11409155527750652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,3584,0.2747306558820936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,3072,0.22472090191311309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,3584,0.11491821871863471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,2560,0.18931733237372506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,3072,0.08777955505583022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,3584,0.24239110946655273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,2048,0.16327910953097874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,2560,0.07648355431026883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,2048,0.06574577755398221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,2560,0.2426053418053521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,2048,0.15171021885342068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,1536,0.05205333232879639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,768,0.05945510996712578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,1536,0.1272782219780816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,1024,0.04218577676349216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,512,0.05494577685991923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,768,0.04092622134420607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,768,0.06995199786292182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,256,0.033506668276256986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,512,0.038430223862330117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,512,0.06088888645172119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,128,0.024080889092551336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,4096,0.28473244773017037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,256,0.032784889141718544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,64,0.020584889584117465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,256,0.05488711264398363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,32,0.020610666937298246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,3072,0.2049413389629788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,128,0.0513591104083591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,64,0.051326221889919706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,1024,0.0779928896162245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,32,0.05130399929152595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,51200,1024,0.11797332763671875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,65536,0.5545902252197266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,51200,128,0.03158222304450141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,51200,1.0558942159016926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,51200,1536,0.09619200229644775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,51200,0.44622490141126847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,12288,0.23589600457085502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,16384,0.1337653398513794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,51200,1.2182043923272026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,65536,1.6300622092352972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,16384,0.35921867688496906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,12288,0.3077804512447781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,10240,0.27915912204318577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,10240,0.08532000250286526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,8192,0.2266453372107612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,10240,0.22783290015326607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,7168,0.1935288906097412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,8192,0.0692231125301785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,8192,0.25254665480719674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,6144,0.1688151094648573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,7168,0.07321244478225708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,7168,0.15709600183698866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,5120,0.14167466428544787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,6144,0.05747377872467041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,16384,0.3874231179555257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,6144,0.1592586702770657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,4096,0.11331289344363743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,5120,0.053745776414871216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,5120,0.1141857835981581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,3584,0.09999110963609482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,12288,0.10390222072601318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,3072,0.08306133084826998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,4096,0.09318933222028945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,4096,0.040752000278896756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,2560,0.07035466697480944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,3584,0.0844293302959866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,3072,0.03323555654949612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,2048,0.0577404432826572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,3072,0.07182311349444918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,2560,0.02852977646721734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,1536,0.039472887913386025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,2560,0.06227911180920071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,2048,0.024711999628278945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,1024,0.028419554233551025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,65536,1.3378782272338867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,1536,0.020111110475328233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,2048,0.0525146656566196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,768,0.02212000058756934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,1536,0.04140533341301812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,1024,0.01719911065366533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,512,0.016159110599093966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,768,0.015844444433848064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,256,0.01204355557759603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,512,0.014475555883513557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,768,0.02625333269437154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,128,0.008905777500735389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,512,0.021925333473417494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,256,0.013162666724787818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,64,0.00790666706032223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,16384,32,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,256,0.019554666346973844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,128,0.013153778182135688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,128,0.018582221534517076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,16384,3584,0.0351893338892195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,64,0.01885511146651374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,32,0.01888799998495314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,16384,0.27143467797173393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,51200,0.3302968872918023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,65536,0.42058756616380477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,51200,0.710661358303494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,16384,1024,0.03159288896454705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,65536,1.0229298273722331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,12288,0.2040728860431247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,16384,0.2764497862921821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,16384,0.10445066955354478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,12288,0.08140266603893705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,10240,0.17539644241333008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,51200,0.9143501917521158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,8192,0.13867910703023276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,10240,0.06864355670081244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,65536,1.21670044793023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,12288,0.21902222103542754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,10240,0.18678311506907144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,6144,0.10562044382095337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,8192,0.13716267214881048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,7168,0.04989688926272922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,5120,0.09131644169489543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,7168,0.12360178099738227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,6144,0.04502400093608432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,4096,0.07336888710657756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,6144,0.10638488663567437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,5120,0.038001779052946306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,3584,0.0648053354687161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,5120,0.08944533268610637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,4096,0.03204266561402215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,3072,0.05309066507551405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,4096,0.07340888844596015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,2560,0.050866666767332286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,3584,0.06406844324535794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,3072,0.025823111335436504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,2048,0.0365555551317003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,3072,0.056215114063686795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,2560,0.02367555598417918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,1536,0.029606223106384277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,2560,0.0494026674164666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,1024,0.021143111917707656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,2048,0.04100711146990458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,1536,0.017147555947303772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,768,0.01624000072479248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,1536,0.03386488887998793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,1024,0.014607111612955729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,8192,0.06439199712541369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,1024,0.025455999705526564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,512,0.012881778180599213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,768,0.013106666505336761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,256,0.009280888570679558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,768,0.021222222182485793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,512,0.0122079998254776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,128,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,512,0.017497777938842773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,256,0.011437333292431302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,64,0.006313777632183499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,256,0.015800888339678448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,3584,0.029437333345413208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,32,0.0063271112740039825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,128,0.01143022212717268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,128,0.015146666102939181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,64,0.01515111161602868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,12288,32,0.015269334117571512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,12288,2048,0.020053333706325956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,12288,7168,0.10565600130293105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,65536,0.3486657672458225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,65536,0.7863910992940267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,16384,0.19484710693359375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,51200,0.6840222146775988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,51200,0.2823839982350667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,12288,0.1470844480726454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,16384,0.09820622205734253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,65536,1.0293235778808594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,12288,0.07343733310699463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,51200,0.7711777687072754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,12288,0.18728444311353895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,8192,0.14304444524976942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,10240,0.05961155229144626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,7168,0.12006133132510716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,10240,0.1492560042275323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,8192,0.11808799372778998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,8192,0.04983733428849114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,6144,0.09096266825993855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,5120,0.09126755264070298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,7168,0.10559822453392877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,7168,0.049141334162818066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,6144,0.03715199894375271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,4096,0.0664186676343282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,6144,0.08968089024225871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,5120,0.033131556378470525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,3584,0.05388889047834608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,4096,0.027613333529896204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,5120,0.0758462217119005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,3072,0.04644711150063408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,2560,0.04126133190260993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,3584,0.02542488939232296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,3072,0.022970666488011677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,3584,0.05524888965818617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,2048,0.03469689024819268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,3072,0.04844533403714498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,2560,0.02103999919361538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,16384,0.23907110426161024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,1536,0.02667466633849674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,10240,0.1524346669514974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,2560,0.042408890194363065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,1024,0.018457776970333524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,2048,0.018353778454992507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,768,0.01447111037042406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,2048,0.035516444179746844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,1024,0.013467555244763693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,512,0.011401777466138205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,1536,0.028480887413024902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,256,0.007720888488822513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,768,0.012141333685980903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,1024,0.022111111217074927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,768,0.01888799998495314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,512,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,512,0.015526221858130561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,64,0.005658666706747479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,256,0.013801777528391944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,32,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,128,0.013536888692114087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,64,0.01347644461525811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,4096,0.06389333142174615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,10240,32,0.013735111388895245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,1536,0.015781333049138386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,65536,0.7191093232896594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,65536,0.30880623393588597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,51200,0.5828728675842285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,256,0.010658666491508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,10240,128,0.010695111420419482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,65536,0.799415111541748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,51200,0.26053955819871694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,12288,0.13621155420939127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,16384,0.21133155292934844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,10240,0.09931911362542047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,51200,0.6109662055969238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,12288,0.14737866984473333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,12288,0.06279644701215956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,8192,0.0957217746310764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,10240,0.049765331877602466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,8192,0.0418524444103241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,7168,0.08926933341556126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,8192,0.0966764423582289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,7168,0.036502222220102944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,6144,0.07070489062203301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,7168,0.08532977766460842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,10240,128,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,5120,0.06426578097873263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,6144,0.032347556617524885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,16384,0.194961773024665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,4096,0.04803555541568332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,6144,0.07552978065278795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,5120,0.028127110666698877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,3584,0.04488977789878845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,5120,0.06266399886873034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,4096,0.024021334118313257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,16384,0.07469955417844984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,3072,0.03364888827006022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,3584,0.022151110900772944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,4096,0.05087200138303968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,3584,0.04558666547139486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,2560,0.02927910950448778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,3072,0.0199955552816391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,2048,0.027031110392676458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,1536,0.02125511070092519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,2560,0.01829688913292355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,3072,0.040707555082109236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,1024,0.015513777732849121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,2048,0.01613333324591319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,2560,0.03535822365019057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,10240,0.1222791141933865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,1536,0.01369599997997284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,2048,0.029682666063308716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,512,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,1024,0.011915555430783166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,768,0.011047110789351992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,256,0.007455999652544658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,512,0.01036711119943195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,768,0.01591822173860338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,128,0.006508444332414203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,512,0.013461333182122974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,256,0.009709333380063375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,64,0.005776888794369168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,256,0.012413333687517377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,8192,128,0.009758222434255812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,32,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,128,0.011744889120260874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,64,0.011749332977665795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,32,0.011711999773979187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,65536,0.4866737789577908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,8192,768,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,65536,0.2538693216111925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,51200,0.4407057762145996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,16384,0.14372177918752035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,1024,0.018806222412321303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,12288,0.1113742192586263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,65536,0.7135333485073514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,51200,0.19985244009229872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,16384,0.0726977785428365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,51200,0.5411991013420953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,10240,0.10676711135440403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,16384,0.16895022657182482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,12288,0.12520799371931288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,10240,0.04760977625846863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,7168,0.06433510780334473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,8192,0.03444622291458978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,10240,0.1062320007218255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,6144,0.06273511383268568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,7168,0.031155556440353394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,7168,0.07881066534254286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,5120,0.05008711086379158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,6144,0.02765422231621212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,6144,0.0657164454460144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,4096,0.04187377625041538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,5120,0.024833778540293377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,3584,0.03412088751792908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,5120,0.0562373333507114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,4096,0.021299555897712708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,8192,1536,0.024538666009902954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,3072,0.031493332650926374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,4096,0.04603644543223911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,3584,0.01945955554644267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,2560,0.025506666964954797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,3072,0.01811555524667104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,3584,0.041045334604051374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,2048,0.02108088963561588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,3072,0.03669688767857022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,2560,0.016558221644825406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,12288,0.04999022351370918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,2560,0.031960888041390315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,2048,0.014551111393504672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,8192,0.08158933454089694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,1024,0.011930666863918304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,768,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,1536,0.012669333153300814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,1024,0.010811555716726514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,1536,0.022092445029152766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,8192,0.09578488932715522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,768,0.010238221950001186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,256,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,768,0.014556444353527494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,512,0.009730666875839233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,512,0.0124177775449223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,128,0.005361777626805835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,256,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,256,0.01143200033240848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,64,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,7168,128,0.00904444439543618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,32,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,128,0.011011555790901184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,64,0.011015111373530494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,1536,0.016846223009957206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,32,0.01092711091041565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,2048,0.02699022160636054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,65536,0.4267608854505751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,7168,512,0.0074986666440963745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,7168,1024,0.016757332616382174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,51200,0.3950088818868001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,65536,0.23628534211052787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,16384,0.13166756100124782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,51200,0.1875386635462443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,16384,0.06054311328464084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,65536,0.6092160012986925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,10240,0.09339911407894558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,16384,0.14675288730197483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,51200,0.48223468992445206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,12288,0.04937066634496053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,12288,0.11096444394853379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,8192,0.06612444586224027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,7168,0.06383022334840563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,10240,0.04316177633073595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,8192,0.03387644555833604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,10240,0.09292711151970758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,6144,0.05880888965394762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,8192,0.0765191117922465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,5120,0.044399112462997437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,7168,0.030155556069480047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,6144,0.02700977855258518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,7168,0.06737066639794244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,4096,0.03614666726854112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,3584,0.031215111414591473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,5120,0.02402666707833608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,4096,0.020631111330456205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,5120,0.05020266771316528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,3072,0.029130667448043823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,3584,0.019152889649073284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,2560,0.02438755498992072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,3584,0.037535998556349016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,3072,0.01744888888465034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,2048,0.020215999748971727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,12288,0.10109688838322957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,2560,0.01572533282968733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,1536,0.01572800013754103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,2560,0.029348442951838177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,2048,0.014137778017255994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,1024,0.011395555403497485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,2048,0.02461244497034285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,768,0.009367110828558603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,1536,0.012333333492279053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,1024,0.010860444770918952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,1536,0.01997244523631202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,512,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,1024,0.015547555353906421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,768,0.010000000397364298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,256,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,768,0.013456888496875763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,128,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,256,0.008987555901209513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,512,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,6144,0.05927466683917575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,64,0.004393777913517422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,256,0.010437332921557956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,6144,32,0.004551110996140374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,128,0.009033777647548253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,128,0.010022222167915767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,4096,0.041441778341929116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,64,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,32,0.010055111514197456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,6144,3072,0.03352266550064087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,65536,0.4711102379692926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,16384,0.12098222308688694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,51200,0.17506844467586943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,65536,0.5223209063212076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,51200,0.40067021052042645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,6144,512,0.009338666995366415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,12288,0.04398222102059258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,12288,0.09665689203474258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,10240,0.07852355639139812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,8192,0.06290666924582587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,10240,0.038541333542929754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,10240,0.08089511262045966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,8192,0.030809776650534734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,8192,0.06675910949707031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,51200,0.3305644459194607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,65536,0.2230142222510444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,7168,0.02714311083157857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,6144,0.04965333475006951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,5120,0.040582223070992365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,6144,0.025571554899215698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,12288,0.0921048919359843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,4096,0.03288088904486762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,6144,0.05145688851674398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,16384,0.138446225060357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,5120,0.021914665897687275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,5120,0.044138666656282216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,3584,0.02860977914598253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,4096,0.01866133345497979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,3072,0.024300444457266066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,4096,0.03642755415704515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,3584,0.017076444294717576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,2560,0.020572443803151447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,3584,0.032976888948016696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,3072,0.015807110402319167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,2048,0.0173342227935791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,3072,0.02991022335158454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,2560,0.014284445179833306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,7168,0.056582219070858426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,2560,0.026179555389616225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,2048,0.01276177747382058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,1024,0.00980977796845966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,2048,0.0221742226017846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,1536,0.011036444041464063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,768,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,1536,0.017895110779338412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,7168,0.059148444069756397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,1024,0.009543110926946005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,512,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,16384,0.05629422267278036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,1024,0.013896889156765409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,768,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,512,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,768,0.012400888734393649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,128,0.0047315553658538395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,512,0.010450666977299584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,64,0.0041759999261962045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,256,0.009700444837411245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,32,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,128,0.009118222528033787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,64,0.009307555854320526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,5120,32,0.009242666264375051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,1536,0.013946667313575745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,65536,0.38990399572584367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,65536,0.17366488774617514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,51200,0.267124440934923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,65536,0.4239653216467963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,16384,0.09352711174223159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,5120,256,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,51200,0.13780800501505533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,12288,0.07122488816579182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,51200,0.351473781797621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,16384,0.046794666184319385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,10240,0.06253244479497273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,16384,0.10635111067030166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,128,0.00850933293501536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,12288,0.037661333878835045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,12288,0.08098755280176799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,10240,0.03292266527811686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,7168,0.04396888944837782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,8192,0.026576888230111863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,7168,0.024340444140964087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,8192,0.05579644441604614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,7168,0.05031377739376492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,5120,0.029839999145931665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,6144,0.022084444761276245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,6144,0.04370311233732435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,4096,0.026644445127911035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,5120,0.01980799933274587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,3584,0.02221333318286472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,5120,0.0380711125002967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,4096,0.016928889685206942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,3072,0.020208888583713107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,4096,0.031566222508748375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,3584,0.015446222490734525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,2560,0.018222222725550335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,3584,0.028247111373477515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,3072,0.014488000008794995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,5120,256,0.008100444244013892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,2048,0.014736889137162102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,2560,0.013064000341627332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,8192,0.04927822285228305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,1536,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,2560,0.022447999980714586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,2048,0.0120666664507654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,1024,0.00924177798959944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,6144,0.03841244512134128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,1536,0.010673777924643623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,1024,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,1536,0.015480889214409722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,1024,0.012392889294359418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,768,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,768,0.011012444065676795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,256,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,512,0.008407110969225565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,512,0.009461333354314169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,128,0.0042088888585567474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,256,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,256,0.008639111287064022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,4096,128,0.007958222594526079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,64,0.003863999827040566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,32,0.003899555653333664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,128,0.008329778081840938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,3072,0.02573866645495097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,64,0.00830844458606508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,32,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,10240,0.06855733527077569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,768,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,4096,512,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,65536,0.14111199643876818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,16384,0.08032533195283678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,65536,0.3875111209021674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,51200,0.11095555623372395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,51200,0.2968506548139784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,12288,0.060023109118143715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,16384,0.03703644540574815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,10240,0.05104710989528232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,16384,0.09596800141864353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,12288,0.02955377764172024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,8192,0.040251556369993426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,12288,0.07263288895289104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,10240,0.025962667332755193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,7168,0.034140444464153714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,10240,0.06242933538224962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,8192,0.021567111214001972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,65536,0.2579093244340685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,6144,0.030844443374209937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,8192,0.05113333463668823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,4096,2048,0.01886488828394148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,5120,0.026211554805437725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,7168,0.045365333557128906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,4096,0.02054755555258857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,5120,0.016303999556435477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,51200,0.2539706759982639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,6144,0.039832890033721924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,5120,0.03420533405409919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,3584,0.01848355597919888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,3072,0.01687022215790219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,4096,0.01443822185198466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,3584,0.01311377767059538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,4096,0.02867199977238973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,3584,0.026173333326975506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,2560,0.014512888259357877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,3072,0.012057777908113269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,2048,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,3072,0.023349333140585158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,1536,0.010013333625263637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,2560,0.01108266661564509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,2048,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,2560,0.01995377739270528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,1024,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,2048,0.017636444833543565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,1536,0.009102221992280748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,768,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,1536,0.014463999205165438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,1024,0.00834222220712238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,512,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,768,0.008017777568764156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,1024,0.011396444506115384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,256,0.0042053332759274375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,7168,0.019871999820073444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,768,0.010316444767846001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,128,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,512,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,64,0.0031617778456873367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,6144,0.018039110634062026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,256,0.007248000138335758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3584,128,0.007358222372002072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,256,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,128,0.007679111427730984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,64,0.007763555480374231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,65536,0.22694222132364908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,32,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,16384,0.06879111131032307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,51200,0.22121777799394396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,65536,0.13360267215304905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,12288,0.05330666568544176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,51200,0.1059911118613349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,16384,0.03645955522855123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,65536,0.34511378076341415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,10240,0.04504088891877068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,51200,0.26507555113898384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,16384,0.0844773319032457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,12288,0.029141333368089464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,8192,0.03580088747872247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,7168,0.03089955449104309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,10240,0.025952888859642878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,12288,0.06513777706358168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,6144,0.026919111609458923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,8192,0.021274665991465252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,8192,0.04550844430923462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,5120,0.021846221552954778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,6144,0.01797155539194743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,7168,0.04087200098567539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3584,512,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,4096,0.01850666602452596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,6144,0.03607022100024753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,3584,0.0166631109184689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,4096,0.01389600005414751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,4096,0.025973333252800837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,3072,0.014310222533014087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,3584,0.012949333422713809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,3584,0.023659555448426142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,2560,0.012879999975363413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,3072,0.011872889267073737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,3072,0.020983111527231004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,2560,0.010868444210953183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,2560,0.018557333283954196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,2048,0.01076533314254549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,10240,0.05516711208555433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,1536,0.008800889054934183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,7168,0.019502222537994385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,2048,0.009711999860074785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,1024,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,2048,0.015614221493403116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3584,32,0.003495111233658261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,1536,0.008988444175985124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,768,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,1536,0.013232889274756113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,5120,0.015954666667514376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,512,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,1024,0.008343111309740279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,1024,0.010676444404655032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,768,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,512,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,768,0.009673777553770278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,128,0.0035297779573334586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,256,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,512,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,64,0.003162666741344664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,32,0.003169777699642711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,256,0.007733333441946242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,3072,128,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,128,0.007735110819339752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,64,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,65536,0.20283022191789415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,32,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,51200,0.1603777805964152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,65536,0.1310008896721734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,16384,0.06194755766126844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,65536,0.29530488120185006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,51200,0.10453155305650498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,12288,0.0435751113626692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,51200,0.2280933327145047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,16384,0.03579377796914842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,16384,0.0750560031996833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,10240,0.03879822293917338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,12288,0.02903999884923299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,8192,0.031543999910354614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,3072,5120,0.03111999895837572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,10240,0.025632888078689575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,7168,0.028216001060273912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,8192,0.020669332808918424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,10240,0.04859911070929634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,6144,0.024413333998786077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,8192,0.0403439998626709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,5120,0.021359999974568684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,7168,0.03665688965055678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,6144,0.017305778132544625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,4096,0.017717333303557504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,6144,0.03187377916442023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,5120,0.015365334020720588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,3584,0.0158906661801868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,5120,0.027878221538331773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,4096,0.013461333182122974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,3072,0.014045332868893942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,3584,0.01278133359220293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,4096,0.023288889063729182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,2560,0.012485333614879184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,3072,0.011848889291286469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,3584,0.020624000165197585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,2048,0.01014666673209932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,3072,0.018820444742838543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,2560,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,1536,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,2048,0.009703999592198266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,2560,0.016743110285864938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,1024,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,1536,0.008973333570692275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,2048,0.01421688828203413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,768,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,12288,0.056531555122799344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,1536,0.012025777664449481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,512,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,1024,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,768,0.00776622196038564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,256,0.0038924444880750445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,768,0.00869599978129069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,512,0.007337777978844113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,128,0.0032248888164758682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,512,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,256,0.007080000307824876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,64,0.0031600000543726813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2560,32,0.0031519999934567344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,128,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,128,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,64,0.007252444823582967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,65536,0.20434933238559297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,32,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,51200,0.16524267196655273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,65536,0.12684799565209284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,65536,0.24643201298183867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,16384,0.052582220898734204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,51200,0.11174310578240289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,51200,0.1949146721098158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,3072,256,0.0038986665507157645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,1024,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,12288,0.03789688812361823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,16384,0.03369333346684774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,10240,0.038231998682022095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,16384,0.06338400310940213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,12288,0.027671999401516382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,8192,0.024453333682484094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2560,7168,0.019191111127535503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,10240,0.024822221861945257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,7168,0.022534221410751343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,12288,0.04881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,10240,0.04194400045606825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,6144,0.01924000018172794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,8192,0.01999022232161628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2560,256,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,5120,0.017099555995729234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,7168,0.018277333842383493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,4096,0.014195554786258273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,6144,0.016754666964213055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,7168,0.031694223483403526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,3584,0.013393777940008374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,5120,0.01479200025399526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,5120,0.024307555622524683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,4096,0.013321778012646569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,3072,0.011420444481902652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,3584,0.012053333222866058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,4096,0.020106666617923312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,2560,0.00998133338159985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,3584,0.01846933364868164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,2048,0.008388444781303406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,3072,0.01641244524055057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,2560,0.0103493332862854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,1536,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,2560,0.014508444401952954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,2048,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,2048,0.012665777570671506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,1536,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,1536,0.010748444332016839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,768,0.005008888741334279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,1024,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,1024,0.008950221869680617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,512,0.0042168887125121225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,768,0.007624000310897827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,8192,0.03506844573550754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,256,0.0035182221068276297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,768,0.008003555238246918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,512,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,128,0.0028106667515304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,6144,0.02784444557295905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,512,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,256,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,32,0.0028239999794297745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,256,0.00666933341158761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,128,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,3072,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,64,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,65536,0.1712364488177829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,2048,32,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,51200,0.13262311617533365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,65536,0.08018222120073107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,16384,0.04621866676542494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,1024,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,65536,0.20331199963887533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,12288,0.03498933381504483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,16384,0.02306488818592495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,10240,0.028563555743959215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,16384,0.05316355493333605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,12288,0.019567110472255282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,8192,0.02292711039384206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,12288,0.04115733173158433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,10240,0.017800889081425138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,7168,0.02032355633046892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,10240,0.03532622257868449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,2048,64,0.002809777855873108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,8192,0.029725333054860432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,2048,128,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,7168,0.013287111288971372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,5120,0.015432000160217285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,7168,0.02718577782313029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,6144,0.012413333687517377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,4096,0.012987555729018318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,6144,0.023697778582572937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,5120,0.011129777464601727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,3584,0.01203733351495531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,5120,0.020572443803151447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,4096,0.010242666635248397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,3072,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,51200,0.06526133086946276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,3584,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,2560,0.00926755534278022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,3584,0.015768888923856948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,51200,0.16129867235819498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,3072,0.009413333402739631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,2048,0.00775022225247489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,3072,0.014173333843549093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,1536,0.0063777777055899305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,2048,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,2048,0.01107377807299296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,8192,0.014474666780895658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,1536,0.007873777714040544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,1024,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,1536,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,768,0.004568888909286923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,6144,0.01777866648303138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,1024,0.007290666302045186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,1024,0.00797155582242542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,512,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,768,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,256,0.0031519999934567344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,768,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,512,0.006606222026877933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,128,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,512,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,256,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,64,0.0025600000388092464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,4096,0.016999999682108562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1536,32,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,128,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,128,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,65536,0.11019911368687947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,64,0.00598933340774642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,2560,0.012749333348539142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,51200,0.08879999981986152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,65536,0.07972355683644612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,65536,0.1548737817340427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,16384,0.03195110956827799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,51200,0.06429421901702881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,12288,0.032087110810809664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,51200,0.12588444021013048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,16384,0.022435555855433147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,10240,0.028468443287743464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,16384,0.04238222373856438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,12288,0.018698667486508686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,8192,0.02048711147573259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,12288,0.03295200069745382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,10240,0.01684266659948561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,7168,0.02071555621094174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,8192,0.01385244478782018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,256,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,6144,0.01692800058258904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,8192,0.023936889237827722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,5120,0.016360888878504436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,7168,0.02177422245343526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,6144,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,4096,0.009846222069528367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1536,2560,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,6144,0.0196079992585712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,5120,0.010749333434634738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1536,32,0.006246222390068903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,3584,0.008803555534945594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,5120,0.016576889488432143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,3072,0.009712888962692684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,4096,0.009871111147933537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,2560,0.0070053330726093715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,3584,0.009331555830107795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,4096,0.014145778285132514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,3072,0.009027555584907532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,2048,0.006213333457708359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,3584,0.013080889152155982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,1536,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,2560,0.008650666309727563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,2560,0.01075466639465756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,2048,0.00833511104186376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,1024,0.004179555508825514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,1536,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,2048,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,768,0.0038417776425679526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,1536,0.008429333567619324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,1024,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,512,0.003335111050142182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,1024,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,256,0.0027635554886526535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,768,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,512,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,128,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,7168,0.012792888614866467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,512,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,64,0.002452444492114915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,256,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,128,0.0063075555695427795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,256,0.00564177789621883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,128,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,65536,0.09567022323608398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,64,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,32,0.005694222119119432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,51200,0.08017422093285455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,65536,0.07919200261433919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,3072,0.011753777662913004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,16384,0.029779556724760268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,65536,0.1521715588039822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,51200,0.0636319981680976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,12288,0.03006311257680257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,51200,0.12153689066569011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,10240,0.023440000083711412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,16384,0.04100266761249966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,1024,768,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,1024,10240,0.02892177634769016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,10240,0.016813332835833233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,10240,0.028307557106018066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,1024,32,0.002461333241727617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,8192,0.023864888482623633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,7168,0.015613334046469795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,7168,0.012850667039553324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,6144,0.013735999663670858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,7168,0.021591999464564856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,6144,0.011936888926559024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,5120,0.012645333177513547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,6144,0.018978666928079393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,16384,0.02215733296341366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,4096,0.009568000005351173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,5120,0.011047999891969891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,12288,0.018311111463440787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,3584,0.008866666919655269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,5120,0.01662488944000668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,8192,0.018410666121376883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,4096,0.009873777627944946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,3072,0.00776977754301495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,3584,0.009559111462699043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,2560,0.007111111448870764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,3584,0.01291911138428582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,3072,0.008961777720186446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,2048,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,8192,0.013677332964208392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,3072,0.012050666742854647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,1536,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,2560,0.008649778034951951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,2560,0.010822222464614444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,2048,0.008005333443482717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,1024,0.004286222159862518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,1536,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,2048,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,768,0.003822222352027893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,1024,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,1536,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,768,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,1024,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,256,0.002814222127199173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,512,0.006485333459244833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,768,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,128,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,512,0.006228444476922353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,256,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,64,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,32,0.0022391111900409064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,768,128,0.006260444306664997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,12288,0.03269244564904107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,65536,0.06284799840715197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,128,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,64,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,4096,0.014118222726715935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,32,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,65536,0.0794080032242669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,16384,0.02092355489730835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,65536,0.11100888914532131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,51200,0.06329066885842217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,12288,0.017549332645204332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,16384,0.02138311167558034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,51200,0.08762844403584798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,10240,0.015586665934986539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,16384,0.031014220582114324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,12288,0.018174222773975797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,10240,0.0165928883685006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,768,512,0.003185777821474605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,7168,0.011389333340856763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,8192,0.013454222016864352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,8192,0.018185777796639335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,6144,0.010054222411579555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,7168,0.012713777522246042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,7168,0.016573333077960543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,5120,0.008656889200210571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,768,256,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,6144,0.011715555356608497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,4096,0.007546666595670912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,6144,0.015046222342385186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,5120,0.010658666491508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,3584,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,5120,0.013442666994200813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,4096,0.009649777577983009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,3072,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,4096,0.011734222372372946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,3584,0.009307555854320526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,2560,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,3584,0.010770666930410596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,3072,0.00884622252649731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,2048,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,3072,0.009938666390048133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,8192,0.012821333275900947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,1536,0.0041804446114434134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,12288,0.02475022276242574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,2560,0.009323555562231276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,2048,0.00793422261873881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,10240,0.021532444490326777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,2048,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,768,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,1024,0.006979555719428592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,1536,0.007483555210961237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,512,0.0028515555378463534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,1024,0.0063279999627007385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,768,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,768,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,256,0.0025111111915773815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,512,0.006305777778228124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,512,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,128,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,256,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,256,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,64,0.0021840000732077495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,128,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,128,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,32,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,65536,0.03904089000489976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,2560,0.008344888687133789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,64,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,512,32,0.005146666533417172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,1024,0.0034977777136696708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,65536,0.07879555225372314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,16384,0.014582221706708273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,65536,0.10928355322943793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,51200,0.06273777617348565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,12288,0.01237511138121287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,51200,0.08646666341357762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,16384,0.02093422247303857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,10240,0.011021333436171213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,16384,0.030272000365787085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,12288,0.01790577835506863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,512,51200,0.048659556441836886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,8192,0.009770666559537252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,12288,0.024023110667864483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,10240,0.016286222471131217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,10240,0.020827554994159274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,8192,0.013432888521088494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,6144,0.009495110975371467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,8192,0.017788444956143696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,7168,0.012400888734393649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,7168,0.016173332929611206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,5120,0.008273777862389883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,6144,0.01165066659450531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,4096,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,6144,0.014787554740905762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,5120,0.010678222609890832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,51200,0.03329688972897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,3584,0.006302222195598815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,5120,0.01312800000111262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,512,1536,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,4096,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,3072,0.005764444255166584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,2560,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,3584,0.00925244473748737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,3072,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,3584,0.01054666688044866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,2048,0.004519999855094486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,2560,0.0083155557513237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,1536,0.003834666477309333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,3072,0.009714666340086196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,2560,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,1024,0.0031946665710873077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,1536,0.007408888803588019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,2048,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,7168,0.010671999719407825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,768,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,512,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,1024,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,768,0.006649777707126405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,1024,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,256,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,512,0.0063235556913746735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,768,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,128,0.002199999988079071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,512,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,256,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,128,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,256,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,32,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,128,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,65536,0.03194933467441135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,64,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,4096,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,51200,0.026038222842746313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,65536,0.07842044697867499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,16384,0.011692444483439127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,65536,0.10837244325213963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,51200,0.06252977583143446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,256,2048,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,12288,0.009778666827413771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,51200,0.08555822239981757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,16384,0.020999999509917364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,10240,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,16384,0.02949777907795376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,12288,0.017820444371965196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,1536,0.007426666716734569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,10240,0.016195555528004963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,12288,0.0236151119073232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,7168,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,10240,0.020720889170964558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,8192,0.013182222015327878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,6144,0.00700711127784517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,8192,0.017458667357762653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,7168,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,256,64,0.001820444439848264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,5120,0.007322666545708974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,7168,0.016195555528004963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,6144,0.011471110913488599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,4096,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,6144,0.014565333724021912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,5120,0.01054933336046007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,3584,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,5120,0.013110222087966071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,4096,0.009642666412724389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,4096,0.011398221883508893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,3072,0.00563822231358952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,3584,0.009134222235944536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,2560,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,3584,0.010497777826256223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,3072,0.00886311133702596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,2048,0.004498666773239772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,2560,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,3072,0.00979644474056032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,1536,0.0038933331767717996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,2560,0.009144888983832465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,2048,0.007688889073001013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,1024,0.003168888803985384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,1536,0.007479111353556315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,2048,0.00832799987660514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,8192,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,768,0.0029102222373088202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,1024,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,512,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,1536,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,256,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,768,0.006702222343948152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,1024,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,512,0.0064222220745351576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,128,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,768,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,256,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,64,0.0019022222194406723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,512,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,128,32,0.0021679999513758556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1025,128,128,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,256,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,65536,0.030357334348890517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,51200,0.022944000032213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,128,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,64,0.005102222164471944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,128,32,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,12288,0.009400000174840292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,256,32,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,10240,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,65536,0.10785777701271905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,51200,0.08543999989827473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,8192,0.0075084442893664045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,7168,0.0070879997478591064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,16384,0.029431111282772485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,5120,0.00792266676823298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,12288,0.023404444257418316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,10240,0.020294222566816542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,4096,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,8192,0.017337777548366122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,3584,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,7168,0.016062221593326993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,6144,0.014431110686726041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,3072,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,2560,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,5120,0.013067555924256643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,2048,0.0042453333735466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,1536,0.0038515557017591265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,4096,0.011275555524561139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,3072,0.009673777553770278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,1024,0.003140444349911478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,2560,0.009031111167536842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,768,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,512,0.002463999928699599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,2048,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,1536,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,256,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,128,0.0018880000958840053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,1024,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,16384,0.01034222212102678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,64,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,512,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,256,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,128,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,65536,0.02506133251720005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,51200,0.018031999468803406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,64,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,6144,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,32,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,16384,0.016364443633291457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,12288,0.011167111496130625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,10240,0.010461333725187512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,65536,0.10860355695088704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,51200,0.08506221903695001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,16384,0.02920177910063002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,12288,0.023335999912685815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,7168,0.013271110753218332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,6144,0.011596444580290051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,10240,0.02056533263789283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,3584,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,5120,0.010215111076831818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,8192,0.0173973325226042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,4096,0.00850933293501536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,7168,0.016184000505341422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,6144,0.014448889427714877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,3584,0.007325333025720384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,3072,0.006501333167155583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,5120,0.012789333032237159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,2560,0.006066666709052191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,64,768,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,4096,0.011318222516112857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,2048,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,64,32,0.0019422223170598347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,1536,0.004223111189073986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,1024,0.0033235556135574975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,3584,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,768,0.0030942221896515954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,2560,0.00885688927438524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,512,0.002607999990383784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,2048,0.008015111088752747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,1536,0.007323555648326874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,256,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,1024,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,128,0.0021297778520319196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,64,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,768,0.005977777971161737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,32,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1025,32,8192,0.00848355558183458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,512,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,256,0.005283555636803309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,128,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,64,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,32,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1025,32,3072,0.009920888476901585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,16384,1.2515920003255208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,16384,1.4726951387193468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,16384,0.4189235634273953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,12288,0.9420159657796224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,51200,1.37408447265625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,12288,1.0153350830078125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,12288,0.3208835654788547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,10240,1.0253083970811632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,16384,1.186523543463813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,51200,3.9436658223470054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,10240,0.832298649681939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,8192,0.621347533331977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,10240,0.26527200804816353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,51200,4.704010857476129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,8192,0.6526871257358128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,8192,0.22171200646294487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,7168,0.5541324615478516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,10240,0.7148693402608236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,7168,0.1947253280215793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,7168,0.5661679903666178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,8192,0.5871724552578396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,6144,0.5059955384996202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,12288,0.8484364615546333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,6144,0.5175484551323785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,6144,0.17260089185502794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,5120,0.3744995594024658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,4096,0.30412888526916504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,5120,0.14777599440680608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,5120,0.42735910415649414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,7168,0.514410654703776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,4096,0.3495768970913357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,3584,0.2831502225663927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,6144,0.43646756807963055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,5120,0.3704986572265625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,3584,0.30048267046610516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,3072,0.2221484449174669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,3584,0.10675377978218926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,4096,0.29617155922783744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,3072,0.2588275538550483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,2560,0.21355911095937094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,3072,0.09661600324842666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,3584,0.2672293451097277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,2048,0.1600968837738037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,2560,0.08351822031868829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,3072,0.23675200674268934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,2560,0.2135173347261217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,1536,0.1263777812321981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,2560,0.19718044333987764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,2048,0.07359288798438178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,1024,0.08395644691255356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,4096,0.12309332688649495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,1536,0.0577404432826572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,1536,0.13747555679745144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,768,0.06140177779727512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,1024,0.04610310991605123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,1024,0.10529155863655938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,1024,0.10122133625878228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,512,0.052376889520221286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,768,0.041751109891467623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,768,0.08338489135106404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,768,0.08357422219382392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,512,0.07200177510579427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,256,0.03307911091380649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,512,0.07042400042215984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,512,0.03747911254564921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,2048,0.17187466886308458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,128,0.031197332673602637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,256,0.06561155451668634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,2048,0.16308711634741888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,256,0.05492977632416619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,64,0.026132444540659588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,128,0.061066667238871254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,1536,0.1313262250688341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,65536,32,0.02667377723587884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,128,0.03381333417362637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,128,0.042240000433391996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,64,0.0605306691593594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,32,0.06070844332377116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,51200,3.8850869072808156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,256,0.03468088971243964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,65536,1.3743955824110243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,51200,3.2374267578125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,65536,4.073894924587674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,16384,1.3235368728637695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,51200,1.0897529390123155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,51200,3.705304039849175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,65536,4.660622914632161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,16384,0.32216978073120117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,16384,1.078577783372667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,12288,0.9614213307698568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,12288,0.25060976876152885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,12288,0.7818720075819227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,10240,0.6607297791375054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,10240,0.21606844001346162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,8192,0.5646755430433485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,10240,0.7251697646247016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,12288,0.6656924353705512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,16384,0.8878711064656576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,8192,0.1767493353949653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,10240,0.5516755315992568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,8192,0.5123759905497233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,7168,0.4194648795657688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,7168,0.1543226639429728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,6144,0.3347128762139215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,7168,0.47176800833808047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,5120,0.2777457767062717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,6144,0.4428533448113336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,6144,0.1371635595957438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,7168,0.40029067463344997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,51200,3.048719194200304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,5120,0.1161688831117418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,4096,0.2654399871826172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,6144,0.3392808967166477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,3584,0.21371732817755806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,5120,0.29147643513149685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,4096,0.10070577594969009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,4096,0.28180355495876735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,4096,0.23599643177456328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,3584,0.0870328876707289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,3072,0.2040648857752482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,8192,0.4604746500651042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,3584,0.22919021712409127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,65536,3.933149125840929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,3072,0.08145688639746772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,2560,0.17569155163235137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,3072,0.1931742164823744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,2048,0.13645511203342015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,3072,0.18443732791476777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,2560,0.1696071094936795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,5120,0.3331502278645833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,1536,0.09863199790318807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,2048,0.057971556981404625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,2048,0.13610400093926325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,2560,0.1717226637734307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,1024,0.061547557512919106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,2048,0.13199378384484187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,1536,0.04608355628119575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,1536,0.10802221960491604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,768,0.056045333544413246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,1536,0.10618399911456639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,1024,0.03742311067051358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,512,0.04417333337995741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,768,0.03416622347301907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,1024,0.08279644118414985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,768,0.06663555569118924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,1024,0.08152266343434651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,256,0.02693333393997616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,768,0.06781511174307929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,512,0.05773244301478068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,512,0.031199110878838435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,128,0.02219999995496538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,256,0.028727110889222886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,512,0.05737511316935221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,64,0.01982133256064521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,256,0.05258666806750827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,256,0.045012444257736206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,51200,32,0.020709334148301017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,2560,0.06885600090026855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,51200,128,0.02795999911096361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,128,0.04905866583188375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,128,0.03496444556448195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,64,0.048561778333452016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,51200,32,0.04881599876615736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,51200,3584,0.2095093329747518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,65536,0.4673626687791612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,51200,1.0519528918796115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,65536,1.2697537740071614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,16384,0.3641351064046224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,51200,1.187321768866645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,65536,1.4793031480577257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,51200,0.3731244405110677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,16384,0.3594800101386176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,16384,0.1136195527182685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,12288,0.259368896484375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,12288,0.08789066473642985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,16384,0.2971911165449354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,10240,0.2167431116104126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,12288,0.22408888075086805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,51200,0.9798622131347656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,10240,0.21051643954383004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,10240,0.07551644245783488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,8192,0.17221511734856498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,10240,0.18878666559855142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,7168,0.13312444421980116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,65536,1.3003546396891277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,8192,0.0611413319905599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,8192,0.18808089362250435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,6144,0.1445351176791721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,12288,0.24609867731730142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,8192,0.15257510874006483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,7168,0.1472782161500719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,7168,0.058168888092041016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,5120,0.11407466729482015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,6144,0.04716622167163425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,7168,0.14270667235056558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,4096,0.08991822269227769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,5120,0.04157244496875339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,5120,0.10851555400424534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,5120,0.10476799805959065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,6144,0.11902133623758952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,3584,0.07744977871576945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,4096,0.03525155451562669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,4096,0.09057866864734226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,3072,0.07204177644517687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,3584,0.03087644444571601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,4096,0.08703466918733384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,3584,0.07928444279564752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,2560,0.05724089013205635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,3584,0.07648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,3072,0.027855998939938013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,3072,0.0685280031628079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,2048,0.04750666684574551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,3072,0.07023377550972833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,2560,0.02569688856601715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,1536,0.03546844588385688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,2560,0.05959733327229818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,2048,0.02202044427394867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,2560,0.061075554953681103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,1024,0.02464177707831065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,2048,0.04935555656750997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,1536,0.018464888135592144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,2048,0.052128000391854175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,768,0.0188391109307607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,1536,0.03930044505331252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,6144,0.13371466265784368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,1024,0.015605333778593274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,512,0.014138667119873894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,1024,0.030223111311594646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,768,0.014070222775141398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,1024,0.034719109535217285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,256,0.010678222609890832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,768,0.02522755497031742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,512,0.012786666552225748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,768,0.02972888946533203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,512,0.020969778299331665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,128,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,256,0.011968889170222811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,512,0.02568888829814063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,64,0.007684444387753804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,256,0.01868266695075565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,16384,32,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,128,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,256,0.02084355552991231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,128,0.01793511046303643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,128,0.016935111747847665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,64,0.01794755624400245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,32,0.017907554904619854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,1536,0.043824000491036304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,65536,0.3654044469197591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,65536,0.8831395573086209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,51200,0.7713600264655219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,16384,0.2985573344760471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,51200,0.8680711322360568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,51200,0.28368979030185276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,16384,0.10438666741053264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,16384,0.27376000086466473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,12288,0.1971262163586087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,16384,0.26054133309258354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,10240,0.17058044009738496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,12288,0.0710115565194024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,12288,0.2027742200427585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,8192,0.1454942226409912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,10240,0.060255997710757785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,12288,0.191703995068868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,10240,0.15945422649383545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,51200,0.7732702361212836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,10240,0.17672444714440239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,8192,0.04929600159327189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,8192,0.13219200240241155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,65536,1.0449724197387695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,8192,0.1304524474673801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,65536,1.0733697679307725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,7168,0.04379822148217095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,7168,0.11547111140357123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,5120,0.08526400062772964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,7168,0.11657688352796768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,6144,0.038853334056006536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,6144,0.09927111201816136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,4096,0.06705511278576322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,5120,0.03389511174625821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,6144,0.10273422135247125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,5120,0.08449333243899876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,3584,0.05989244249131945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,4096,0.028730667299694482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,4096,0.06833333439297147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,4096,0.07418577538596259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,7168,0.10687910848193699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,3072,0.05410666598214043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,3584,0.026194666822751362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,3584,0.06064711014429728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,3584,0.06690755817625257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,2560,0.0428017775217692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,3072,0.023636443747414485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,6144,0.10495111015107896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,3072,0.05398844348059761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,2048,0.03490755624241299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,2560,0.021470222208235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,2560,0.04682844546106127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,1536,0.026576888230111863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,2048,0.018811555372344125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,2048,0.03885866536034478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,2048,0.04402933186954922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,1536,0.031457778480317854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,1024,0.01846222248342302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,1536,0.016398222910033334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,5120,0.08779555559158325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,1536,0.03794399897257487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,768,0.015413333972295126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,1024,0.013696889082590738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,1024,0.02437155610985226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,512,0.011733333269755045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,768,0.012215110990736218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,768,0.026591999663247004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,512,0.016851555969980028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,512,0.011411555939250521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,256,0.011258666714032492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,3072,0.059976001580556236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,512,0.02223199937078688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,2560,0.051670223474502563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,256,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,256,0.015156444576051501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,128,0.010875555376211802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,64,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,32,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,128,0.015332443846596612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,64,0.01479200025399526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,32,0.014547554983033074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,768,0.020710221595234342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,1024,0.029690666331185236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,12288,128,0.00943288869327969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,65536,0.9370329115125867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,65536,0.31020977762010365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,256,0.018018666240904067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,51200,0.6236035558912489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,65536,0.908773316277398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,16384,0.21674933698442247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,51200,0.24600177341037324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,51200,0.7181182437472873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,16384,0.22456712192959258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,16384,0.21220178074306914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,12288,0.18076711230807832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,12288,0.05947822332382202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,10240,0.15139467186397976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,65536,0.8606017960442437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,12288,0.16201510694291857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,128,0.01482399966981676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,10240,0.057895110713111035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,8192,0.1094684468375312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,10240,0.134479112095303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,10240,0.14254044161902532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,8192,0.04293955696953667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,7168,0.10437155432171291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,51200,0.6624577840169271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,8192,0.10949066612455581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,6144,0.09759910901387532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,8192,0.11799111631181504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,7168,0.0376835564772288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,16384,0.07897155814700656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,5120,0.08234844605127971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,7168,0.10159289174609715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,6144,0.03414933217896356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,7168,0.10063911146587795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,6144,0.08462222417195638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,4096,0.06586844391292997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,5120,0.02945599953333537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,6144,0.09410311116112603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,3584,0.05686666568120321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,5120,0.07174488570955065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,12288,0.16736533906724718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,4096,0.02539288832081689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,4096,0.059302224053276904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,5120,0.07781510882907443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,3072,0.04831377665201823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,4096,0.06457866562737359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,3584,0.02274666726589203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,2560,0.038217776351504855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,3584,0.0528400010532803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,3072,0.02087200019094679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,3584,0.05783110857009888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,3072,0.04674666788842943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,2560,0.019132445255915325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,3072,0.05239466826121012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,2560,0.040714667903052434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,1536,0.023271110322740342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,2048,0.016872000363137987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,2048,0.033815112378862165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,1024,0.01640977793269687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,1536,0.013958222336239286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,1536,0.027671999401516382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,1536,0.0332551101843516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,768,0.013434666726324292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,1024,0.012039999994966718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,1024,0.026927111877335444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,768,0.01816177699300978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,512,0.010355555348926121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,768,0.011394666300879585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,768,0.023574221465322707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,2048,0.032433778047561646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,512,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,256,0.0074648890230390764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,512,0.01516711049609714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,512,0.02050044470363193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,128,0.006375999914275275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,256,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,2560,0.0459555553065406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,256,0.013439999686347114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,64,0.007440889047251807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,10240,32,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,256,0.01718933383623759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,128,0.00980711148844825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,128,0.01309511148267322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,2048,0.03907733493381076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,128,0.013911111487282647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,64,0.01312444441848331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,32,0.013074666261672974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,1024,0.021412443783548143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,51200,0.4512017567952474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,65536,0.25602844026353627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,65536,0.676238218943278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,51200,0.19874133004082573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,16384,0.15745688809288874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,65536,0.7417519887288412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,51200,0.582223097483317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,16384,0.06435822116004096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,16384,0.18445689148373076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,10240,0.11247911718156602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,16384,0.21003022458818224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,12288,0.13446311155954996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,12288,0.13410311275058323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,51200,0.5342773331536187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,8192,0.09302400218115912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,10240,0.043466667334238686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,65536,0.6818781958685981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,10240,0.11955022811889648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,7168,0.0826551119486491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,8192,0.035199112362331815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,8192,0.09451733032862346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,6144,0.06757599777645536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,8192,0.0958462225066291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,7168,0.03186400069130792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,7168,0.08078222142325507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,7168,0.09634933206770156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,6144,0.028628445333904688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,12288,0.12011110782623291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,6144,0.07125244537989299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,6144,0.07572799921035767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,5120,0.024745777249336243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,4096,0.045808000697029956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,12288,0.050204445918401085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,5120,0.06167111131880018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,3584,0.04145066605673896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,5120,0.06569955746332805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,4096,0.04900000161594815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,3584,0.044567998912599355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,3584,0.019924443629052904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,3584,0.04946488804287381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,3072,0.01789422167672051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,3072,0.03847733471128676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,2560,0.030681778987248737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,3072,0.04502755403518677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,5120,0.057962669266594775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,2560,0.016632889707883198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,2048,0.02500266664557987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,2560,0.038776887787712946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,2048,0.028561777538723413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,2048,0.014706666270891825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,1536,0.01944266590807173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,2048,0.0339955555068122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,4096,0.02166844407717387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,3072,0.03485600153605143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,4096,0.054808888170454234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,1024,0.014129777749379476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,1536,0.023444443941116333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,1536,0.028931554820802476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,768,0.011726222104496427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,1024,0.01088355564408832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,1024,0.017895999881956313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,1024,0.023583110835817125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,512,0.008889777792824639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,768,0.010101333260536194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,768,0.015483554866578845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,768,0.020543111695183646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,2560,0.03376977642377218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,256,0.006320000108745363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,512,0.009739555418491364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,512,0.013112000293201871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,128,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,512,0.018022222651375663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,256,0.009038222332795462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,256,0.011969777444998423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,64,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,256,0.0148026661740409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,8192,32,0.006486222147941589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,128,0.011556444068749746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,1536,0.012803555362754397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,10240,0.11163200272454156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,64,0.011414222419261932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,32,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,65536,0.5900426440768772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,65536,0.23521600829230416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,51200,0.46354044808281797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,16384,0.15338488419850668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,65536,0.6543217764960395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,51200,0.18571466869778105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,51200,0.5047057999504937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,128,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,12288,0.11269599861568874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,16384,0.05920000208748711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,128,0.012820444173283048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,16384,0.1565893358654446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,16384,0.17539732986026338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,10240,0.10163466797934638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,12288,0.04689066608746847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,12288,0.11876000298394097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,8192,0.08981511327955459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,10240,0.04024799995952182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,10240,0.10061866707271999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,51200,0.5299030939737955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,7168,0.07899110847049289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,10240,0.1131582260131836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,8192,0.08191644483142428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,6144,0.06551377640830146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,8192,0.09433777464760675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,7168,0.029857777886920508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,7168,0.0735786689652337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,5120,0.055550220939848155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,7168,0.08481511142518784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,6144,0.026924444569481745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,6144,0.06336444616317749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,4096,0.0439262224568261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,6144,0.07401866383022733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,5120,0.023991111252042983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,5120,0.0546248886320326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,3584,0.03968088825543722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,5120,0.06433600187301636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,4096,0.0207342223988639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,4096,0.04450133442878723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,65536,0.6882710986667209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,3072,0.033823112646738686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,4096,0.053598221805360585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,3584,0.019055111540688407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,12288,0.13415111435784235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,2560,0.029559112257427637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,3584,0.04042311178313361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,3072,0.01770400007565816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,3584,0.049103998475604586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,3072,0.036107556687461004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,8192,0.03302222159173753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,2560,0.015862221519152325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,3072,0.04326488905482822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,1536,0.019256000717480976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,2560,0.03130844566557143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,2560,0.038262221548292376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,1024,0.014522666732470194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,2048,0.026591999663247004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,1536,0.01257155587275823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,2048,0.03316266669167413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,1536,0.021584888299306233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,768,0.012148444851239523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,1536,0.027865777413050335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,1024,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,1024,0.016770665844281513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,1024,0.023010666171709698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,768,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,768,0.014230221509933472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,256,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,768,0.020248888267411124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,512,0.009316444396972656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,512,0.012053333222866058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,512,0.017301332619455125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,128,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,256,0.011015111373530494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,256,0.014514666464593677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,2048,0.02467022173934513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,64,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,128,0.010686222049925061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,2048,0.01384800010257297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,32,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,128,0.012010667059156628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,64,0.010755555497275459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,32,0.010663111176755695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,7168,512,0.008794666992293464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,65536,0.4622631072998047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,65536,0.20132177405887178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,51200,0.33907021416558164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,65536,0.5610195795694987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,256,0.00869688888390859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,16384,0.1120693286259969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,51200,0.4475777943929036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,128,0.008799999952316284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,16384,0.13603022363450792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,12288,0.10019022226333618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,65536,0.5366302066379124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,16384,0.14113333490159777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,12288,0.04320177767011854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,10240,0.08004888561036852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,12288,0.10404711299472386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,51200,0.42167290051778156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,12288,0.1095306674639384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,8192,0.07957866456773546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,10240,0.03525777657826742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,10240,0.0975102252430386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,10240,0.09290044175253974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,8192,0.029081778393851385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,8192,0.0731146666738722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,6144,0.055938667721218534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,8192,0.08027288648817274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,7168,0.02595377796226078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,7168,0.06507466899024116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,5120,0.045850667688581676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,51200,0.17595911026000977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,7168,0.06949955224990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,6144,0.02400355537732442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,6144,0.0565848880343967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,4096,0.03837955660290188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,16384,0.055663108825683594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,6144,0.06126488579644097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,3584,0.0329075555006663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,5120,0.02152266601721446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,4096,0.018426666657129925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,5120,0.05323733223809136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,4096,0.03993155558904012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,3072,0.0269795556863149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,4096,0.04506133331192864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,3584,0.016730666160583496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,3584,0.035614222288131714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,2560,0.024523556232452393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,3584,0.040780444939931236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,3072,0.015464888678656684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,3072,0.032002667586008705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,7168,0.0687191088994344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,2560,0.014126222994592456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,3072,0.036554667684766985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,2560,0.027928001350826685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,1536,0.014195554786258273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,2560,0.03321689036157396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,2048,0.012979555461141797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,2048,0.023598222268952265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,1024,0.011592888997660743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,2048,0.028543998797734577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,1536,0.011074666347768573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,768,0.009738666315873465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,1536,0.024658666716681585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,1024,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,1024,0.014954666296641031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,512,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,1024,0.0199644449684355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,768,0.009017777939637503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,5120,0.048134220971001514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,768,0.013112000293201871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,256,0.006474666711356904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,768,0.017852443787786696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,512,0.00847822262181176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,512,0.011031111081441244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,128,0.0058444444504049085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,512,0.015807999504937064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,256,0.010101333260536194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,256,0.00812177773978975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,64,0.00535733335547977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,32,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,6144,2048,0.02015199926164415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,128,0.009731555150614845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,128,0.011714666253990598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,64,0.00998488896422916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,32,0.009742221898502773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,1536,0.019262222780121695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,65536,0.4615102344089084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,65536,0.18161243862575957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,51200,0.32839467790391713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,65536,0.4908622105916341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,16384,0.11852444542778863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,51200,0.14271199703216553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,51200,0.4056364430321588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,256,0.014581332604090372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,65536,0.5281306902567545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,128,0.008020444048775567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,16384,0.1197155581580268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,12288,0.0794524417983161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,16384,0.05041955577002632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,10240,0.058911111619737416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,12288,0.0384204453892178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,51200,0.41469422976175946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,16384,0.1423164473639594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,8192,0.061425778600904674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,12288,0.09216266870498657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,12288,0.10784800185097589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,7168,0.05341688791910807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,8192,0.027302222119437322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,8192,0.062344888846079506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,8192,0.07588889201482137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,7168,0.02490844494766659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,6144,0.04513422316975064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,7168,0.05498844385147095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,7168,0.06849866443210177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,5120,0.036304887798097395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,6144,0.022389334109094407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,6144,0.04842399888568454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,4096,0.029088000456492107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,5120,0.020073778099483915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,6144,0.06000088983111911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,5120,0.04162044326464335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,3584,0.026206221845414903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,5120,0.05169511172506544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,4096,0.01761511133776771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,4096,0.0347360008292728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,3072,0.02215555475817786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,3584,0.01612622208065457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,10240,0.033498668008380465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,3584,0.03123733401298523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,2560,0.019208000765906442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,10240,0.09203110800849067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,3072,0.014762666490342883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,3072,0.027983999914593164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,2048,0.01499466598033905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,3072,0.03635199864705404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,2560,0.01334755536582735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,2560,0.024731554918819006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,1536,0.012753778033786349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,2560,0.032418668270111084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,2048,0.02068177858988444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,2048,0.027740443746248882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,1536,0.016985777351591323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,1024,0.009024000002278222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,1536,0.010592888626787396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,1536,0.023781332704755995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,768,0.007798222204049428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,1024,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,1024,0.013242666920026144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,1024,0.01919911139541202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,512,0.006319111006127463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,10240,0.07768888605965509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,768,0.011718221836619906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,3584,0.04025777843263414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,256,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,512,0.008460444708665213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,512,0.010055111514197456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,512,0.015505777464972602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,128,0.004552888787455029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,256,0.009332444104883406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,256,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,2048,0.01205599970287747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,256,0.014482667048772177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,64,0.004402666870090696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,128,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,5120,32,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,128,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,128,0.010959111154079437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,64,0.009034666750166152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,32,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,65536,0.34574045075310605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,4096,0.043931557072533496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,51200,0.22960800594753691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,768,0.008976889153321585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,65536,0.4014631112416585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,16384,0.09104355838563706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,51200,0.11150399843851726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,65536,0.3894933329688178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,16384,0.09812621937857734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,16384,0.03749777873357137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,12288,0.06267555554707845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,16384,0.10668177737130059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,10240,0.05634578069051107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,12288,0.029839999145931665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,12288,0.07508444123797946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,12288,0.08226666847864787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,10240,0.06339555316501193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,10240,0.026293334033754136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,8192,0.045575999551349215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,10240,0.07081866926617093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,65536,0.13909510771433511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,7168,0.04108533263206482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,8192,0.05188266767395867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,768,0.0174666659699546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,8192,0.02162577708562215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,8192,0.05860266420576307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,6144,0.03600800037384033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,51200,0.3226835462782118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,7168,0.020208888583713107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,5120,0.02952622373898824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,7168,0.04628622200753954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,6144,0.018197332819302876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,7168,0.05341777867741055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,6144,0.040236443281173706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,4096,0.024325332707828943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,6144,0.04729155699412028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,51200,0.30744534068637425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,5120,0.016484444340070088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,5120,0.03524622321128845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,3584,0.02180800007449256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,5120,0.0418177776866489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,4096,0.014482667048772177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,4096,0.029168887270821467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,3584,0.013400000002649097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,4096,0.03563288847605387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,3584,0.02643555568324195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,2560,0.01662577854262458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,3584,0.032794667614830866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,3072,0.012146666646003723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,3072,0.023872888750500146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,2048,0.014292443792025248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,3072,0.0292524430486891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,2560,0.011345778074529437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,2560,0.020869332883093093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,1536,0.011958222422334882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,2560,0.02652799917591943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,2048,0.01789866718981001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,1024,0.008901333643330468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,2048,0.02330844435426924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,1536,0.009141333401203156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,1536,0.014668444792429606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,768,0.007179555793603261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,1536,0.019539554913838703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,1024,0.008349333372380998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,1024,0.011682666838169098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,512,0.005872888697518243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,1024,0.016159999701711867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,768,0.008031999899281396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,768,0.010393777655230628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,256,0.004592888885074192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,768,0.015188443991872998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,512,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,512,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,128,0.004221333397759331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,512,0.013831111292044321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,256,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,256,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,64,0.003933333274390963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,32,0.0039022221333450745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,256,0.01275288893116845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,128,0.008363555702898238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,128,0.010011555420027839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,64,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,2048,0.00998488896422916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,32,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,65536,0.3173840045928955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,51200,0.23823110262552896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,65536,0.1330488920211792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,65536,0.3764195442199707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,16384,0.08493866523106892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,51200,0.2911377747853597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,51200,0.10602577527364095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,4096,3072,0.01901155544651879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,12288,0.06512710783216688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,65536,0.38453245162963867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,16384,0.09227910968992446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,16384,0.036815110180113055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,51200,0.305531554751926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,16384,0.10594311025407578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,12288,0.029622223642137315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,12288,0.07129955291748047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,128,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,8192,0.04429866539107429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,10240,0.026085333691702947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,12288,0.08155733346939087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,10240,0.060738669501410596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,7168,0.03871022330390082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,10240,0.07100888755586413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,8192,0.02164799968401591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,6144,0.03388444582621256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,8192,0.04940977692604065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,7168,0.019876443677478366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,5120,0.029050668080647785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,8192,0.05806666612625122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,6144,0.01834133267402649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,7168,0.052710221873389355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,6144,0.03917155663172404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,4096,0.023038221730126276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,6144,0.047112001313103564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,5120,0.016328000360065036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,5120,0.03400533397992452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,3584,0.021651556094487507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,5120,0.04099555479155646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,4096,0.014242667290899487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,4096,0.028471999698215063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,3072,0.018616888258192275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,4096,0.03523555397987366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,3584,0.013155555559529198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,10240,0.053448888990614146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,3584,0.025681777132882014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,2560,0.01683111157682207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,3584,0.03222044308980306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,3072,0.012082666986518435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,2048,0.014552889598740472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,3072,0.022985777921146814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,2560,0.011023110813564725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,3072,0.02907555633121067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,2560,0.019959110352728102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,1536,0.011980444192886353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,2560,0.026222222381167944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,2048,0.00999111102686988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,2048,0.01718222267097897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,1024,0.009071110851234859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,7168,0.044343110587861806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,1536,0.009030222064918941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,1536,0.014118222726715935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,768,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,1536,0.01941155559486813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,1024,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,512,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,1024,0.011428444749779172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,1024,0.016199111938476562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,768,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,256,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,768,0.015322667029168872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,512,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,512,0.008957333034939235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,512,0.013775111072593264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,128,0.004206222378545337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,256,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,256,0.00815466625822915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,128,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,128,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,256,0.012089778151777057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,128,0.010260444548394945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,64,0.007684444387753804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,32,0.007828444242477417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,65536,0.2897875573900011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,2048,0.02305511136849721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,51200,0.22016089492373994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,65536,0.3308115535312229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,65536,0.128584000799391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,768,0.01000533335738712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,16384,0.07486577828725179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,51200,0.2600711186726888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,51200,0.10218844148847793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,65536,0.3834639920128717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,12288,0.05545599924193489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,16384,0.0359280010064443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,16384,0.08352710803349812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,64,0.0038373333712418876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,10240,0.04648800028695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,16384,0.10491110881169637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3584,32,0.004184000194072723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,51200,0.32324088944329155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,12288,0.02868888775507609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,12288,0.06398933463626437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,8192,0.03802400165134006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,7168,0.03246577911906772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,10240,0.02570044497648875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,12288,0.08130933178795709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,10240,0.05401777889993456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,8192,0.02129777769247691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,6144,0.028295109669367473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,8192,0.045936889118618436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,10240,0.070269332991706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,7168,0.019367999500698514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,5120,0.023669333921538457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,7168,0.04035022192531162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,6144,0.01789155602455139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,7168,0.052347554100884326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,6144,0.03562044435077243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,5120,0.01586577792962392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,5120,0.03117777903874715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,3584,0.018185777796639335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,5120,0.04081866807407803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,4096,0.013577777478430005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,4096,0.025989333788553875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,3072,0.015560888581805758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,4096,0.034887999296188354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,3584,0.012720000412729053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,3584,0.023568888505299885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,2560,0.013878222141000958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,3584,0.03213511241806878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,3072,0.011709333293967776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,3072,0.020976889464590285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,2048,0.01167022188504537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,3072,0.028917332490285236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,2560,0.010685332947307162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,2560,0.018269333574506972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,2560,0.025056888659795124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,1536,0.009024000002278222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,2048,0.009687110781669617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,2048,0.015842666228612263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,2048,0.02234755622016059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,4096,0.020298666424221463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,1024,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,1536,0.01313777764638265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,6144,0.046410666571723096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,768,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,1536,0.01959022217326694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,1024,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,1024,0.010456889039940305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,512,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,768,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,768,0.00959288908375634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,768,0.015092444088723926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,256,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,512,0.007326222128338284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,512,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,512,0.013697778185208639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,128,0.0035617777870761026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,256,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,256,0.012400888734393649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,8192,0.057638221316867404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,64,0.0031724443866146933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,128,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,128,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,3072,32,0.0035013332962989807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,128,0.01035911093155543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,1536,0.008985777695973715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,64,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,65536,0.23044888178507486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,32,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,1024,0.01614488826857673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,51200,0.17953688568539092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,65536,0.12670221593644884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,65536,0.2868151134914822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,16384,0.059431109163496226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,51200,0.22116622659895155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,51200,0.10242399904463027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,12288,0.043418665726979576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,256,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,16384,0.07333066728379992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,16384,0.033951110310024686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,10240,0.03397333290841844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,51200,0.3016364574432373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,12288,0.027609777119424608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,8192,0.029829333225886028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,12288,0.05899288919236925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,10240,0.024734222226672705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,12288,0.08050222529305352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,7168,0.02568888829814063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,10240,0.0687111086315579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,8192,0.02057422200838725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,8192,0.03973599937227037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,6144,0.021923555268181696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,8192,0.05737688806321886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,7168,0.018605333235528734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,7168,0.03565244542227851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,5120,0.01884622209601932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,7168,0.051656001143985324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,6144,0.016927111479971144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,6144,0.031423999203575984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,4096,0.01565511193540361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,6144,0.04565866788228353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,65536,0.38358932071261936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,5120,0.014898666077189975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,5120,0.027553778555658128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,3584,0.014077333940400017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,5120,0.04000799854596456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,4096,0.01309333327743742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,4096,0.023000889354281958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,3072,0.012512000070677863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,4096,0.03393955694304572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,3584,0.012434666355450949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,3584,0.020625778370433383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,2560,0.01093066649304496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,3584,0.03119200136926439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,3072,0.011422221859296164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,10240,0.0478942228688134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,2048,0.009335999687512716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,3072,0.018391999933454726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,3072,0.028229332632488672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,2560,0.016024000114864774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,1536,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,2048,0.009529777699046666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,2560,0.025270221961869135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,1024,0.005983110931184556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,2048,0.014096000128322177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,2048,0.021273778544531927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,1536,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,768,0.005205333232879639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,1536,0.018904000520706177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,1024,0.008023111356629265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,1024,0.015842666228612263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,512,0.00452622233165635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,768,0.007660444411966536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,768,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,768,0.014708444476127625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,16384,0.10434044731987847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,512,0.007898666792445714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,512,0.007418666448858049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,512,0.013436444103717804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,2560,0.010462221999963125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,256,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,128,0.0032257777121331957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,256,0.012242666549152799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,128,0.0070222218831380205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,128,0.006908444480763541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,64,0.0032035555276605818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,32,0.003212444484233856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,128,0.009986666341622671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,1024,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,65536,0.1916639937294854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,64,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,32,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,51200,0.14639111359914145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,65536,0.0828862190246582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2560,256,0.003543111185232798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,65536,0.24553065829806855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,51200,0.06634310881296794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,256,0.007100444700982835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,16384,0.04999555481804741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,51200,0.18932978312174478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,12288,0.036970665057500206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,51200,0.19381154908074272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,16384,0.023689778314696416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,1536,0.00870311094654931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,16384,0.06306044260660808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,10240,0.031395554542541504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,12288,0.019559999306996662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,16384,0.07095466719733344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,12288,0.048397332429885864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,8192,0.024495111571417913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,12288,0.05519644419352213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,10240,0.017997332745128207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,10240,0.041492445601357356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,7168,0.02273511058754391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,10240,0.048047112094031445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,8192,0.01479466590616438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,8192,0.03465866711404588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,8192,0.040163556734720864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,7168,0.013564444250530668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,65536,0.24470578299628365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,5120,0.016740444633695815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,6144,0.01257511145538754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,7168,0.036785778072145246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,6144,0.027768888407283362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,4096,0.014007111390431723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,6144,0.032856888241238065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,5120,0.011382222175598145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,5120,0.024315555890401203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,3584,0.012994666894276937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,5120,0.02903644575013055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,4096,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,4096,0.02564711040920681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,3584,0.009770666559537252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,3584,0.018405333161354065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,3584,0.02351466649108463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,2560,0.010008888940016428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,3072,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,3072,0.016546666622161865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,3072,0.02071911096572876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,2560,0.008858666651778752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,2048,0.008379555410808986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,2560,0.014479999740918478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,6144,0.019467555814319186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,2560,0.01887022289964888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,1536,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,7168,0.03124799993303087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,2048,0.008286221987671321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,1024,0.00553688903649648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,2048,0.012713777522246042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,2048,0.01687733332316081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,1536,0.007742221984598372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,1536,0.010845333337783813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,1536,0.015100444356600443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,1024,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,1024,0.013117333253224691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,512,0.003899555653333664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,768,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,768,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,768,0.012413333687517377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,256,0.0033084443873829315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,512,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,512,0.011787555283970304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,256,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,256,0.006341333190600078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,128,0.0028773333049482773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,256,0.010679110884666443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,128,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,64,0.0028488888508743713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,128,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,32,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,1024,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,4096,0.019880889190567862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,64,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,3072,0.010959111154079437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,65536,0.15856444835662842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,32,0.006568000134494569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,512,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,65536,0.0788631108072069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,65536,0.2002497778998481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,65536,0.241704887813992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,51200,0.1589848862753974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,51200,0.06351377566655476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,16384,0.036533332533306546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,51200,0.18883377975887725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,128,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,12288,0.029500444730122883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,16384,0.022867555419603985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,2048,768,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,16384,0.052689777480231396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,10240,0.02476177778508928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,16384,0.06874133480919732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,12288,0.019327999817000497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,8192,0.019223110543357003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,12288,0.040766222609413996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,10240,0.017603556315104168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,10240,0.03515733281771342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,12288,0.05395644572046068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,7168,0.018537777993414137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,10240,0.0471093323495653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,8192,0.014496889379289416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,8192,0.029345777299669053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,6144,0.01568800045384301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,8192,0.03923822111553616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,7168,0.013122667041089801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,7168,0.026685333914226953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,5120,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,7168,0.03559289044804043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,6144,0.012051555845472546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,6144,0.023590222001075745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,4096,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,6144,0.032270222902297974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,5120,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,5120,0.020230222079488967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,3584,0.010371555884679159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,5120,0.02751022246148851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,4096,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,3072,0.00945688866906696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,4096,0.023973332511054143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,3584,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,3584,0.015517334143320719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,2560,0.008289777570300633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,3584,0.022154667311244543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,3072,0.009078222016493479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,3072,0.01421866648726993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,2048,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,3072,0.020337777005301583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,2560,0.008618666893906063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,2560,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,1536,0.006174222048785951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,51200,0.1187102264828152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,2048,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,2048,0.011122666299343109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,1024,0.005047111047638787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,1536,0.007631111476156447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,2048,0.01645866698688931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,1536,0.009663110805882348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,768,0.004495111190610462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,1536,0.014766222900814481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,1024,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,512,0.003863111138343811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,1024,0.012847111456924014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,768,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,768,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,256,0.0032133333798911837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,512,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,4096,0.016953777935769822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,512,0.006656888872385025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,128,0.0028897778441508612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,512,0.011134222149848938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,256,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,256,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,64,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,256,0.01036266651418474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,128,0.006303999986913469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1536,32,0.0029102222373088202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,128,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,65536,0.09290755457348293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,64,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,2560,0.01829155617290073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,32,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,51200,0.08491377698050605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,65536,0.15467554993099636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,1024,0.00797155582242542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,16384,0.03226844469706217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,51200,0.12236444155375163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,65536,0.2277146710289849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,768,0.012455110748608908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,51200,0.1807840002907647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,16384,0.022156443860795762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,16384,0.04212266537878248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,16384,0.06479199727376302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,12288,0.018592000007629395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,10240,0.01993600030740102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,12288,0.03308444552951389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,12288,0.05111289024353027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,128,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,8192,0.01722666621208191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,10240,0.016521778371598985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,10240,0.028538667493396338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,10240,0.04445866743723551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,7168,0.015192889504962497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,65536,0.07816266351275973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,8192,0.02377066678471035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,8192,0.03723911113209195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,51200,0.06269155608283149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,6144,0.01349955548842748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,7168,0.012729778057999082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,7168,0.02198222279548645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,5120,0.011663999822404651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,7168,0.03382666574584113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,12288,0.024009777439965144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,6144,0.011703111231327057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,4096,0.010204444328943888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,6144,0.01907111042075687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,5120,0.010991111397743225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,6144,0.03044711218939887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,5120,0.01663555536005232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,3584,0.009325332939624786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,5120,0.02739199995994568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,4096,0.014161777165200977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,3072,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,4096,0.023005333211686876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,3584,0.009329777624871995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,3584,0.021293333835071985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,2560,0.007238222493065729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,3072,0.008987555901209513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,3072,0.01182577759027481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,8192,0.013901333014170328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,3072,0.01942755613062117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,2048,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,2560,0.00846400029129452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,2560,0.010936000280910067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,1536,0.005538666827811136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,2560,0.01780177818404304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,2048,0.008021333151393467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,2048,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,1024,0.004243555582231945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,1536,0.007631111476156447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,1536,0.014450665977266101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,1024,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,4096,0.009754666851626502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,1024,0.00703022215101454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,1024,0.012671111358536614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,3584,0.013108444710572561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,512,0.0031973332580592898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,768,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,768,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,768,0.01237777786122428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,256,0.0028035555862718155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,512,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,512,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,128,0.002500444443689452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,512,0.011365333365069495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,256,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,256,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,64,0.0023608888602919048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,256,0.010384000009960597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,128,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,32,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,128,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,2048,0.015849777393870883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,128,0.00812266684240765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,65536,0.08070844411849976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,64,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,51200,0.0649777783287896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,32,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,65536,0.07795466979344685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,1024,768,0.0038480001191298165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,16384,0.025055110454559326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,65536,0.1522595617506239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,51200,0.062381333774990506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,65536,0.2263706790076362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,12288,0.019072888625992667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,16384,0.02176800039079454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,16384,0.040892445378833346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,51200,0.1797928942574395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,10240,0.016121778223249648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,16384,0.06481777959399752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,12288,0.03244266576237149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,8192,0.013053333593739403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,12288,0.050717333952585854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,10240,0.01665866706106398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,10240,0.028370665179358587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,7168,0.012467555701732635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,10240,0.04366933306058248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,8192,0.01347555551264021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,1536,0.00853777759604984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,8192,0.02348444362481435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,6144,0.010880000061459012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,7168,0.012471111284361945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,8192,0.03600000010596381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,5120,0.00944266633854972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,7168,0.02125866711139679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,7168,0.03368977705637614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,6144,0.011768889096048145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,6144,0.018583110637134977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,4096,0.007800000409285228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,6144,0.029375112719006006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,5120,0.010695999695195092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,5120,0.016536888149049547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,3584,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,51200,0.11967289447784424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,4096,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,5120,0.02608355548646715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,4096,0.014058666096793281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,12288,0.017892445127169292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,4096,0.022384888596004907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,3584,0.009315555294354757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,3584,0.013122667041089801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,2560,0.00620888877246115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,3584,0.020971554848882887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,3072,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,3072,0.012054222325483957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,2560,0.008391111261314815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,2560,0.01106844428512785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,2560,0.017391110459963482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,1536,0.004581333448489507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,2048,0.009662222531106737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,2048,0.015449777245521545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,1536,0.008301333420806462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,1024,0.003858666867017746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,1536,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,1536,0.014384888940387301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,768,0.0035386664999855887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,3072,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,1024,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,1024,0.012095111111799875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,512,0.002883555574549569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,768,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,768,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,2048,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,768,0.01202311118443807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,256,0.002584888910253843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,512,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,3072,0.018992889258596633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,512,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,128,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,512,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,256,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,2048,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,256,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,64,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,256,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,128,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,128,0.0052879999081293745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,65536,0.05412711037529839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,128,0.007744889292452071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,64,0.005305777821275923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,51200,0.046210666497548424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,768,32,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,65536,0.077638222111596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,65536,0.11095199982325236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,16384,0.017312000195185345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,1024,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,51200,0.0625680022769504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,65536,0.22439644071790907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,51200,0.08759733041127522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,12288,0.0173688895172543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,16384,0.021599110629823472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,10240,0.015432000160217285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,51200,0.1781617800394694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,16384,0.03115733464558919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,12288,0.017900443739361234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,16384,0.06332800123426649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,8192,0.01314933349688848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,12288,0.02496444516711765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,10240,0.016216889023780823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,12288,0.04998222324583265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,7168,0.011861333416567909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,10240,0.021431111627154883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,8192,0.013415999710559845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,10240,0.04275644487804837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,8192,0.01790844400723775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,7168,0.012677333421177335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,8192,0.035391112168629967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,768,32,0.002471111093958219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,7168,0.01685333251953125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,7168,0.03286844491958618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,6144,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,6144,0.014906666345066495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,6144,0.028727110889222886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,5120,0.010693333215183683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,5120,0.013467555244763693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,3584,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,5120,0.02552266584502326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,4096,0.009664889011118147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,4096,0.011503111157152386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,3072,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,4096,0.022397332721286353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,3584,0.00925511121749878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,3584,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,2560,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,3584,0.02074044446150462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,3072,0.008737777670224508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,3072,0.009740444521109263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,3072,0.018811555372344125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,2560,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,2560,0.009375111096435124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,6144,0.010216889282067617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,2560,0.017172444197866652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,5120,0.008893333375453949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,2048,0.007917332980367873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,2048,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,4096,0.007592889169851939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,2048,0.015474667151769003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,1024,0.0034977777136696708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,1536,0.007352888584136963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,1536,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,768,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,1536,0.013793778088357715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,1024,0.006343110981914733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,1024,0.006973333656787872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,512,0.002834666727317704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,1024,0.012056888805495368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,768,0.006773333168692059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,768,0.006055111272467508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,256,0.002536888958679305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,768,0.011365333365069495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,512,0.0063733334342638654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,512,0.005666666560702854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,128,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,512,0.011085333095656501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,256,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,2048,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,256,0.005304000029961268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,64,0.002120888895458645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,32,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,256,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,128,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,128,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,65536,0.036972443262736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,512,1536,0.004253333227501975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,51200,0.02940533227390713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,64,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,512,32,0.005192888693677055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,65536,0.07696000072691175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,16384,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,65536,0.1091546681192186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,51200,0.06181510951783922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,51200,0.08616177903281318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,65536,0.22319200303819445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,12288,0.01069155583779017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,16384,0.020561777883105807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,51200,0.17780089378356934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,16384,0.030287110143237646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,10240,0.011108444796668159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,16384,0.06297866503397624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,12288,0.017482666505707633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,12288,0.023707555400000677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,12288,0.04925955666436089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,10240,0.016159110599093966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,10240,0.020644444558355544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,7168,0.01091288857989841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,10240,0.0417137775156233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,8192,0.013285333083735572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,8192,0.017806222041447956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,8192,0.03484888871510824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,7168,0.01620444489849938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,6144,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,7168,0.012400888734393649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,128,0.008006222546100616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,5120,0.008390222158696916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,7168,0.03186577889654372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,4096,0.007257777783605788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,6144,0.011504888534545898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,5120,0.010397333237859938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,6144,0.0147733340660731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,3584,0.0063973334100511335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,5120,0.013139555851618448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,5120,0.025407110651334126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,4096,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,3072,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,4096,0.02200177808602651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,3584,0.008977777428097194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,3584,0.010666666759385003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,2560,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,3584,0.020595555504163105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,3072,0.0086666668454806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,8192,0.012025777664449481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,3072,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,2048,0.004520888957712385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,3072,0.018499554859267343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,2560,0.00832799987660514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,1536,0.003846222327815162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,2560,0.009022222624884712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,2560,0.017280000779363845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,2048,0.008242666721343994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,2048,0.00793599999613232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,1024,0.0032106666929192017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,2048,0.015096889601813423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,1536,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,1536,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,768,0.0029173334025674393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,1024,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,1536,0.013805333111021252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,1024,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,512,0.002528888897763358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,1024,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,768,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,768,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,4096,0.009666666388511658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,256,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,768,0.011394666300879585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,512,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,128,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,512,0.010485333700974783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,256,0.006221333311663733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,256,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,64,0.0019120000716712738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,256,0.010013333625263637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,128,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,256,32,0.0021768889079491296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,128,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,65536,0.029495110114415485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,128,0.007621333003044128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,64,0.004964444372389051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,51200,0.0240000006225374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,32,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,65536,0.07654577493667603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,65536,0.10861600107616848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,16384,0.010096888575288985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,65536,0.22385244899325898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,51200,0.060807108879089355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,51200,0.08556799756156074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,12288,0.01204177737236023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,6144,0.02849066588613722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,16384,0.02971200148264567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,16384,0.020616888999938965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,10240,0.010695111420419482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,256,512,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,12288,0.01757600075668759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,16384,0.06183377901713053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,12288,0.023659555448426142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,8192,0.009156444834338294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,12288,0.04823288983768887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,10240,0.016181333197487723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,10240,0.02067466742462582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,7168,0.008655110994974772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,10240,0.041402667760849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,8192,0.013111111190583972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,8192,0.01753155555990007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,6144,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,8192,0.03496977686882019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,7168,0.012433778080675336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,7168,0.016229333149062265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,5120,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,7168,0.031769778993394636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,6144,0.014627556006113688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,4096,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,6144,0.02830844455295139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,5120,0.01314311143424776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,3584,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,5120,0.02502311103873783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,4096,0.009583111438486312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,4096,0.011311111350854238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,3072,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,4096,0.021939555803934734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,3584,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,3584,0.010441777606805166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,2560,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,3584,0.020290666156344943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,3072,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,3072,0.009666666388511658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,2048,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,3072,0.018745778335465323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,2560,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,2560,0.009019555317031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,1536,0.003924444317817688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,2560,0.016873778568373788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,2048,0.007873777714040544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,2048,0.008275555239783393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,1024,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,2048,0.015110222829712762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,6144,0.011456888582971362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,1536,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,1536,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,768,0.0028968888024489083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,1536,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,1024,0.007076444725195567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,5120,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,1024,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,512,0.002557333351837264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,768,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,768,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,1024,0.012070222033394707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,256,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,51200,0.17786399523417154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,512,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,128,0.0021262222694026097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,512,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,256,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,512,0.010114666488435533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,64,0.0019271110908852683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,256,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,128,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,128,32,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,256,0.0103493332862854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,65536,0.029683556821611192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,51200,0.023010666171709698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,64,0.004960888789759742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,16384,0.009341333475377824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,32,0.005060444275538127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,12288,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,65536,0.10833688577016194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,10240,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,51200,0.08521688646740383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,8192,0.00721155520942476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,16384,0.029520889123280842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,7168,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,12288,0.02342133389578925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,6144,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,10240,0.020470221837361652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,5120,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,8192,0.01739644507567088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,4096,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,7168,0.016048888365427654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,3584,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,6144,0.014427555931939019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,3072,0.005538666827811136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,5120,0.012998222476906247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,2560,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,4096,0.011178666518794166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,128,128,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,2048,0.004214222232500712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,128,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,3072,0.009663110805882348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,1024,0.0031439999325407874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,2560,0.009020444419648912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,768,0.0028080000645584534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,2048,0.00814577771557702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,512,0.002529777793420686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,1536,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,256,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,1024,0.006246222390068903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,128,0.002024000096652243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,768,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,64,0.0018053334206342697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,512,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,32,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,256,0.005195555587609609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,65536,0.025311110748185053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,128,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,64,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,768,0.011058666639857821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,32,0.004975111120276981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,16384,0.01073333372672399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,12288,0.0090506664580769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,65536,0.10783111386828953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,10240,0.008271111382378472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,64,3584,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,64,1536,0.0038444445365005066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,8192,0.007674666742483775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,51200,0.08576711018880208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,7168,0.007039111521508958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,6144,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,12288,0.023354666100607976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,5120,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,4096,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,10240,0.02032977839310964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,8192,0.01734755602147844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,7168,0.0159599996275372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,6144,0.014456000592973499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,3072,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,2560,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,5120,0.012768000364303589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,4096,0.011349333657158745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,2048,0.00508266646001074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,1536,0.004221333397759331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,3584,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,1024,0.00350222239891688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,2560,0.008796444369686974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,768,0.002871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,2048,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,512,0.002487999904486868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,1536,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,51200,0.020018666982650757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,256,0.0022159999029503926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,128,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,1024,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,64,0.0018648888087934917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,768,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,32,0.001875555556681421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,256,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,128,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,16384,0.02944444285498725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,64,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,32,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1024,32,3584,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,3072,0.0096133334769143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,16384,1.00910578833686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,16384,0.402965333726671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1024,32,512,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,12288,0.7079226705763074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,16384,1.1180231306287978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,12288,0.30280976825290257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,10240,0.6359342469109429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,12288,0.8323662016126844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,10240,0.6729404661390516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,10240,0.25625867313808864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,51200,1.3329430686102974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,8192,0.5227893193562826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,8192,0.2115937736299303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,7168,0.4558604558308919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,8192,0.5345164404975044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,51200,3.1598871019151478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,6144,0.377654234568278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,7168,0.1830737723244561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,6144,0.16201155715518528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,5120,0.33700354894002277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,6144,0.4213591151767307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,4096,0.2921902338663737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,5120,0.13821066750420466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,4096,0.1161537832683987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,51200,3.8721796671549478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,5120,0.361791107389662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,3584,0.10113688972261216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,3584,0.25086667802598744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,3072,0.08892444769541423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,2560,0.19256356027391222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,4096,0.2743004428015815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,3072,0.19998221927218965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,2048,0.1428097751405504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,2560,0.0778871112399631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,2560,0.1734915574391683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,1536,0.11928710672590469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,2048,0.06702222426732381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,7168,0.531617800394694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,1024,0.0799457761976454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,2048,0.14031733406914607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,768,0.05915377537409464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,1536,0.11092266771528457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,1024,0.0841733349694146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,768,0.04229155513975355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,768,0.0677093333668179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,3584,0.21832088629404703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,256,0.02921688887808058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,512,0.03605511122279697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,128,0.02404444416364034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,256,0.03249333302179972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,256,0.05395822061432732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,64,0.020288000504175823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,128,0.03204266561402215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,128,0.05167644553714328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,32,0.020779555042584736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,64,0.04987200101216634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,32,0.04998044504059685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,1536,0.05224533213509453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,65536,1024,0.04103999998834398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,512,0.04323200053638882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,65536,3072,0.2023093303044637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,65536,512,0.05898577637142605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,16384,0.7726888656616211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,16384,0.8891803953382703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,16384,0.3196951018439399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,51200,0.9950275421142578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,12288,0.5956595738728842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,65536,1.3995982276068792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,51200,2.624569786919488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,65536,3.3460044860839844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,10240,0.49886131286621094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,12288,0.2740737862057156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,51200,2.9433964623345266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,12288,0.6225031216939291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,10240,0.20282222165001762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,8192,0.36849154366387266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,8192,0.16783911652035186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,7168,0.33802045716179746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,65536,3.7136433919270835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,7168,0.36009422938028973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,7168,0.14798222647772893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,5120,0.24219022856818306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,6144,0.12455822361840142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,5120,0.10981066359413995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,6144,0.27592889467875165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,4096,0.18761866622500947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,5120,0.26032177607218426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,3584,0.16959288385179308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,4096,0.09003021981981064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,4096,0.2150426705678304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,3072,0.14100978109571669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,3584,0.08203822374343872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,2560,0.11898399723900689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,10240,0.6320106718275282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,3072,0.15735377205742732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,2048,0.11779556009504531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,3584,0.18668444951375326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,2560,0.13494843906826443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,8192,0.4737422201368544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,2560,0.06328177452087402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,2048,0.05271644393603007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,1024,0.05972177452511258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,2048,0.11141155825720893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,1536,0.04221866528193156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,768,0.04627644353442722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,1024,0.0340266658200158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,6144,0.3346666759914822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,512,0.038218667109807335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,1024,0.0677164461877611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,768,0.031593776411480375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,256,0.024345777100986902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,512,0.029100444581773546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,768,0.05461511015892029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,128,0.021563554803530376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,512,0.04742311106787788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,256,0.02699022160636054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,64,0.018008889423476327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,3072,0.06938755512237549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,32,0.01918577816751268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,128,0.04034488730960422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,51200,1536,0.08805244498782688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,64,0.04007822275161743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,32,0.040082666609022356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,1536,0.08802844418419732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,65536,0.43633778889973956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,65536,1.0499084260728624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,51200,128,0.026531555586391028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,51200,0.8489742279052734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,16384,0.2832791010538737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,65536,1.174097802903917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,16384,0.2730613284640842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,16384,0.11179555786980523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,51200,0.9039982689751519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,12288,0.19217067294650605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,12288,0.08423466814888848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,12288,0.20192178090413412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,10240,0.15346666177113852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,8192,0.1324417723549737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,10240,0.07279466920428805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,7168,0.11037600040435791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,8192,0.05844000313017103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,10240,0.17395555973052979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,6144,0.09526044130325317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,8192,0.13803822464413112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,7168,0.05324622326427036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,5120,0.08090488778220283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,7168,0.12283555666605632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,51200,256,0.04335111048486498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,4096,0.06646400027804904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,6144,0.10512444708082412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,5120,0.04011022382312351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,3584,0.05933688746558296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,5120,0.08938844336403741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,4096,0.03286044465170966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,3072,0.053258667389551796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,3584,0.028705779049131606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,4096,0.07376711236106025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,2560,0.05069422059588962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,3584,0.06502666738298205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,3072,0.026035555534892615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,2048,0.04182844360669454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,3072,0.05651466714011299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,2560,0.023491554790072974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,1536,0.032949331733915545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,1024,0.02328266700108846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,2048,0.0207004447778066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,2560,0.04892888996336195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,768,0.01863466699918111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,1536,0.017865777015686035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,2048,0.040928890307744346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,512,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,1024,0.01499466598033905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,1536,0.032681776417626276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,768,0.013714666995737286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,1024,0.025382222400771245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,256,0.01018311083316803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,128,0.008768888811270395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,512,0.0177848885456721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,256,0.011766222616036734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,64,0.008117333054542542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,256,0.016013332539134555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,128,0.011657777759763928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,16384,32,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,128,0.015499555402331881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,64,0.015489778584904142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,32,0.015507555670208402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,51200,0.34635554419623477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,65536,0.8233715693155924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,65536,0.3412773344251845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,16384,768,0.02143999934196472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,51200,0.6129191186692979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,65536,0.8895058102077908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,6144,0.04506133331192864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,16384,0.22821688652038574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,16384,0.09060088793436687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,12288,0.1388897763358222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,51200,0.7125617663065592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,16384,0.20784710513220894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,10240,0.13701511753929987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,12288,0.0704560014936659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,51200,0.2774159908294678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,8192,0.09851911332872178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,12288,0.15591022703382704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,7168,0.09747644265492757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,10240,0.05848711066775852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,8192,0.05022577775849236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,10240,0.13094133800930446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,6144,0.08481778038872613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,7168,0.04310755597220527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,5120,0.07016088565190633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,6144,0.03656444284651015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,7168,0.09372533029980129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,4096,0.051848000950283475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,6144,0.08305511209699842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,5120,0.032821334070629544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,3584,0.04510133465131124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,4096,0.027098667290475633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,4096,0.05704710880915324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,3072,0.042565332518683545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,3584,0.024549333585633173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,2560,0.03344000048107571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,3584,0.0510755545563168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,3072,0.022105778257052105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,2048,0.029670221938027277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,2560,0.020181333025296528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,2560,0.038337777058283486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,2048,0.017920000685585868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,2048,0.03215377860599094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,8192,0.11418133311801487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,1536,0.015308444698651632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,1024,0.016807110773192514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,1536,0.026144888665941026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,768,0.01353422221210268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,1024,0.013057777451144325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,512,0.010741333166758219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,16384,512,0.01239466667175293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,1024,0.020668443706300523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,768,0.012042666474978128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,5120,0.07119733095169067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,768,0.01755555636352963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,256,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,128,0.006446222050322427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,512,0.011066666907734342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,256,0.01036622209681405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,64,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,512,0.014708444476127625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,32,0.006568000134494569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,12288,128,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,256,0.013289777768982781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,128,0.012720000412729053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,3072,0.04417777723736233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,64,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,12288,32,0.01277599980433782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,12288,1536,0.02180622186925676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,16384,0.1766479942533705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,51200,0.46062665515475804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,65536,0.5790106455485026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,65536,0.28938044442070854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,51200,0.23253512382507324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,16384,0.17679911189609102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,16384,0.07318044371075101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,12288,0.12772977352142334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,12288,0.05917955769432915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,51200,0.56913513607449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,8192,0.08151466978920831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,10240,0.049978666835361056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,7168,0.06877422332763672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,12288,0.1422142187754313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,8192,0.04027022255791558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,10240,0.11195200019412571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,6144,0.06097422043482462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,8192,0.09312533007727729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,7168,0.03667111198107401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,5120,0.05283022258016798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,7168,0.08118310901853773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,6144,0.03187377916442023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,6144,0.07021511263317533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,5120,0.02902844548225403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,5120,0.059968001312679715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,3584,0.036329779360029436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,4096,0.02386755579047733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,3072,0.03147466646300422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,4096,0.04979999860127767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,3584,0.021827555365032617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,2560,0.029198222690158423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,3584,0.04407111141416761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,3072,0.03905511233541701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,2048,0.025791111919615004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,2560,0.018232888645595975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,1536,0.019018666611777413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,2560,0.0339955555068122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,10240,0.11749422550201416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,2048,0.01645866698688931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,65536,0.7501137521531848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,1024,0.013820444544156393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,768,0.011218667030334473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,2048,0.028668443361918133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,1536,0.013809777796268463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,1024,0.01183377785815133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,1536,0.023206222388479445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,768,0.01109155515829722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,1024,0.018358222312397428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,256,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,768,0.015499555402331881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,512,0.010315555665228102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,128,0.0057706667317284485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,4096,0.041857779026031494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,512,0.012963555753231049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,256,0.009682666924264696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,64,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,256,0.011742221812407175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,128,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,128,0.011742221812407175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,64,0.011746666497654386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,10240,32,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,10240,3072,0.019946666227446664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,65536,0.4685315555996365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,51200,0.3673271073235406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,65536,0.23857421345180937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,512,0.00870577742656072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,16384,0.13558044698503283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,65536,0.5991955333285862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,51200,0.46272802352905273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,12288,0.092013332578871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,16384,0.06054666969511244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,51200,0.18723911709255645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,16384,0.14879822731018066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,10240,0.08261244164572822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,10240,32,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,8192,0.07458666960398357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,12288,0.04801511102252536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,12288,0.10961421993043687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,7168,0.06488533152474298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,10240,0.09241511424382527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,6144,0.056702223089006215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,7168,0.03038755390379164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,8192,0.07595377498202853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,5120,0.04778755704561869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,6144,0.026964444253179762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,7168,0.06750577688217163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,4096,0.03973155551486545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,6144,0.05897777610354953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,3584,0.035417778624428645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,5120,0.05038311084111532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,4096,0.02131555477778117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,4096,0.0414044459660848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,3584,0.019142222073343065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,3584,0.03734933336575826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,2560,0.026143110460705225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,3072,0.017504889104101393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,3072,0.03309422069125705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,2048,0.021341333786646526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,2560,0.01612977849112617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,1536,0.017818666166729398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,2560,0.02923822071817186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,2048,0.014257777068350049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,10240,0.041511999236212835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,1024,0.013316444224781461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,2048,0.02438755498992072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,8192,0.033589333295822144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,768,0.010617777705192566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,1024,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,1536,0.020145777199003432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,512,0.008240888516108194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,768,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,1024,0.015604444675975375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,256,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,768,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,5120,0.02407288882467482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,128,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,512,0.011403555671374003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,256,0.008986666798591614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,64,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,256,0.010482666393121084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,3072,0.029863999949561224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,128,0.008738666772842407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,128,0.010334221853150262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,64,0.01016622202263938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,8192,32,0.010132444401582083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,65536,0.45989513397216797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,51200,0.3101573255327013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,65536,0.21975378195444742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,1536,0.012406222522258759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,16384,0.11065244674682617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,65536,0.5408106909857856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,12288,0.08265688684251574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,51200,0.41195376714070636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,51200,0.18780710962083605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,16384,0.05797244442833794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,8192,512,0.009391110804345872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,10240,0.06729066371917725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,16384,0.12796444363064235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,12288,0.09912000099817912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,8192,0.06922844383451673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,12288,0.05035466618008084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,10240,0.03885866536034478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,7168,0.05782488981882731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,8192,32,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,6144,0.04688266581959195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,10240,0.08629688951704238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,7168,0.02990666694111294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,5120,0.034839110241995916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,8192,0.06660977999369304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,6144,0.025199999411900837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,7168,0.05959022045135498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,6144,0.05203555689917671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,5120,0.023810666468408372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,3584,0.029170665476057265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,5120,0.044377777311537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,4096,0.01980177727010515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,3072,0.024881778491867915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,4096,0.03677511215209961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,3584,0.0185244447655148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,2560,0.021215111017227173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,3584,0.03302222159173753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,2048,0.018062222335073683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,2560,0.014896000425020853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,2560,0.025750221477614507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,1536,0.01368711143732071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,2048,0.013400000002649097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,2048,0.022057778305477563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,1024,0.010050666828950247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,1536,0.011922666596041786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,1536,0.018230222993426852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,8192,0.03204800022972955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,768,0.008284444610277811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,1024,0.009949333137936061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,512,0.006526222245560751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,4096,0.030693332354227703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,768,0.009416888985368941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,256,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,768,0.012373333175977072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,512,0.008996444443861643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,512,0.010713777608341642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,128,0.004569777597983678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,256,0.00869599978129069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,64,0.0042204442951414315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,128,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,7168,3072,0.01702844434314304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,7168,32,0.004271999829345279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,3072,0.02932888931698269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,64,0.009697777529557547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,32,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,65536,0.3539271089765761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,65536,0.18551645014021131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,51200,0.36250665452745223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,1024,0.013800889253616333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,65536,0.46331379148695206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,16384,0.12100356154971653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,12288,0.07925333579381307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,51200,0.36046756638420957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,16384,0.049472000863817006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,10240,0.05965511004130045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,16384,0.11241955227322048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,12288,0.04207911094029745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,256,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,8192,0.04766577813360426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,12288,0.11001866393619114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,7168,0.0424177779091729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,7168,128,0.00960711141427358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,8192,0.0277502222193612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,6144,0.041638222005632185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,8192,0.05973333120346069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,7168,0.02494844463136461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,5120,0.03496355480617947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,7168,0.060363557603624135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,6144,0.02233955595228407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,4096,0.028453333510292902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,6144,0.04711911082267761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,5120,0.020097777247428894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,3584,0.02574133376280467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,5120,0.03971288932694329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,3072,0.021328889661365088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,4096,0.03351199958059523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,3584,0.030046221282747056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,3072,0.015086222026083203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,2560,0.01856355534659492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,3072,0.027191110783153113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,2048,0.014863999353514777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,2560,0.013732444081041547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,2560,0.02382933265633053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,1536,0.012777778009573618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,2048,0.012851555314328937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,10240,0.033575110965304904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,2048,0.020250666472646926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,1536,0.011219555305110084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,1536,0.01661066710948944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,768,0.008103999826643202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,1024,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,1024,0.013065777719020844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,512,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,768,0.008967999782827165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,4096,0.01811022228664822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,256,0.005196444276306364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,3584,0.016359110673268635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,512,0.008406221866607666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,512,0.009992889232105678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,128,0.004182222402758068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,256,0.00812355594502555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,64,0.0041724443435668945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,256,0.009006222089131674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,51200,0.16643288400438097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,32,0.004570666700601578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,128,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,64,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,32,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,6144,1024,0.009677333136399588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,65536,0.3057368861304389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,51200,0.2338328891330295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,65536,0.16842310958438447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,16384,0.08974222342173259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,65536,0.4038284354739719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,768,0.011429333024554782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,51200,0.13547466860877141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,12288,0.07169600327809651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,16384,0.04656799965434604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,16384,0.09988622532950507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,10240,0.06030311187108358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,12288,0.03709244396951463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,8192,0.04806488752365112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,12288,0.07628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,10240,0.03219111098183526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,7168,0.04253955682118734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,10240,0.06563288635677762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,8192,0.026183999247021143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,6144,10240,0.07333066728379992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,6144,0.03441333439615037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,7168,0.02397955622937944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,7168,0.047135998805363975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,6144,0.022088888618681166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,6144,0.0409840014245775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,4096,0.0221377776728736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,5120,0.019817777805858187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,5120,0.03564000129699707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,3584,0.021996445126003687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,51200,0.3097306622399224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,4096,0.029903110530641343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,3072,0.01792888840039571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,3584,0.01556799974706438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,6144,128,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,3584,0.027279999521043565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,2560,0.01663555536005232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,3072,0.014153778553009033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,2048,0.013869333598348828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,2560,0.012892444100644855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,3072,0.02423022190729777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,1536,0.011523555550310345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,2560,0.021430222524536982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,8192,0.05315466721852621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,2048,0.012079999678664737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,1536,0.010452444354693094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,2048,0.01825066738658481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,768,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,1536,0.014485332700941296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,1024,0.009353777600659264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,512,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,1024,0.011723555624485016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,256,0.004520888957712385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,512,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,768,0.010398222340477837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,128,0.0038133333954546186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,512,0.009015999734401703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,4096,0.017307556337780423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,256,0.007861333588759104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,64,0.003536888708670934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,256,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,128,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,128,0.008299555215570662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,64,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,5120,32,0.008307555483447181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,65536,0.3092613220214844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,65536,0.13160710864596897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,65536,0.32024089495340985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,5120,0.029601779248979356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,16384,0.06980533070034452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,51200,0.10416089163886176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,51200,0.31941599316067165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,12288,0.058543999989827476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,16384,0.03602577911482917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,16384,0.08181244134902954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,10240,0.04993244343333774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,12288,0.028554668029149372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,32,0.003528888854715559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,12288,0.06933955351511638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,10240,0.025343111819691126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,7168,0.03534577621353997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,8192,0.023025777604844835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,8192,0.044176889790429004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,6144,0.028619554307725694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,7168,0.039176887936062284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,5120,1024,0.008389333056079017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,5120,0.02538577715555827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,6144,0.017553778158293832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,6144,0.03480711248185899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,4096,0.019925332731670804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,5120,0.016083555089102853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,5120,768,0.008656889200210571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,5120,0.029911110798517864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,3584,0.017975999249352347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,4096,0.025835555460717943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,3072,0.015856888559129503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,3584,0.012915555801656512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,8192,0.040906667709350586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,3584,0.023231999741660223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,2560,0.01420177850458357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,3072,0.011707555916574268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,10240,0.05338222119543287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,3072,0.021070222059885662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,2048,0.012114666402339935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,2560,0.010771555205186209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,1536,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,7168,0.019466666711701285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,2048,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,1024,0.007705777883529663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,51200,0.2388950983683268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,1536,0.009717333647939894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,768,0.006384000182151794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,1536,0.0127697777416971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,1024,0.008303999900817871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,512,0.005376889059940974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,1024,0.010409778190983666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,768,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,256,0.004564444637960858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,768,0.009511999785900116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,512,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,4096,0.013753777576817406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,128,0.0038657776183552216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,256,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,64,0.003530666646030214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,256,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,4096,128,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,4096,32,0.003552000141806073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,128,0.007670222057236566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,64,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,65536,0.2058346668879191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,32,0.007663110891977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,2560,0.01790577835506863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,2048,0.015430221954981485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,65536,0.12586843967437744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,16384,0.05929955508973864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,65536,0.28592001067267525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,51200,0.09877866506576538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,12288,0.042371554507149585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,16384,0.03504800134234958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,10240,0.04000622365209792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,16384,0.0728293326165941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,12288,0.02827022141880459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,12288,0.05618844429651896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,8192,0.02870222263866001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,4096,512,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,7168,0.028318223026063707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,10240,0.025097777446111042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,10240,0.047409776184293956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,8192,0.02088088956144121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,6144,0.023965333898862202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,7168,0.019179556104871962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,8192,0.0392853319644928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,5120,0.021358221769332886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,6144,0.018357333209779527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,7168,0.035327110025617815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,4096,0.0173973325226042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,3584,0.014633778068754407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,5120,0.016280000408490498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,5120,0.026872000760502283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,4096,0.013665777941544851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,3072,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,51200,0.16961956024169922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,4096,0.022808000445365906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,3584,0.012994666894276937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,2560,0.011727111207114326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,3584,0.0214044451713562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,2048,0.010344888601038191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,2560,0.011065777805116443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,3072,0.01869066721863217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,1536,0.008743111458089616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,51200,0.22551200124952528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,2560,0.016199111938476562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,1024,0.006514666808976068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,2048,0.013958222336239286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,1536,0.009333333207501305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,768,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,1024,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,1536,0.011412444214026133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,1024,0.009714666340086196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,768,0.008096888661384583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,256,0.0037457777394188773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,768,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,512,0.007592000067234039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,128,0.0031804444475306403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,512,0.00793333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,256,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,64,0.0029528888149393927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,6144,0.030870222383075293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,32,0.0031582222630580268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,128,0.007280888656775157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,64,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,3072,0.01182933317290412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,32,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,65536,0.2532862292395698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,2048,0.009664889011118147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,51200,0.1872968938615587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,65536,0.12211022112104629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,16384,0.06469511323504977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3584,512,0.004570666700601578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,65536,0.2679724428388807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,51200,0.0984862248102824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,51200,0.20781777964697945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,12288,0.0517866677708096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,10240,0.042952001094818115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,16384,0.034599112139807806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,16384,0.0697573357158237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,8192,0.033193777004877724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,12288,0.05320088730918037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,10240,0.024823110964563157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3584,256,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3584,128,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,10240,0.0454675555229187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,6144,0.025655110677083332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,8192,0.020822222034136455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,8192,0.03787555628352695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,7168,0.019495111372735765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,5120,0.021495110458797876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,7168,0.03421422176890903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,6144,0.017088888419999015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,4096,0.017735110388861764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,6144,0.02993511160214742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,5120,0.015678221980730694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,3584,0.016025778320100572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,5120,0.025980444418059453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,4096,0.0140346669488483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,3072,0.014224888549910652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,4096,0.022209778428077698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,3584,0.012404444317022959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,2560,0.012366222010718452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,3072,0.011556444068749746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,2048,0.010586666564146677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,3072,0.018581334087583754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,2560,0.01053244454993142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,1536,0.008529778156015608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,2560,0.01555199921131134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,12288,0.02784622377819485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,2048,0.009844444692134857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,1024,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,2048,0.013411555853154926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,768,0.00553688903649648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,1536,0.011386666860845355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,7168,0.02929333183500502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,1024,0.008232889076073965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,512,0.005029333134492238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,1024,0.009411555197503831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,768,0.007884444461928474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,256,0.004231111043029361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,768,0.00850133349498113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,512,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,128,0.003510222252872255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,512,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,256,0.007183111376232571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,64,0.0032231110251612137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,3072,32,0.003335111050142182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,256,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,128,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,64,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,65536,0.1619982189602322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,3584,0.01993777851263682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,3072,32,0.0069368887278768755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,51200,0.12571556038326687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,65536,0.12001866764492458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,16384,0.04548355605867174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,51200,0.09524622228410508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,65536,0.2288800080617269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,1536,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,16384,0.034790221187803484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,10240,0.029373334513770208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,12288,0.026689777771631878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,16384,0.059783114327324756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,8192,0.0236817780468199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,12288,0.04611555404133267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,7168,0.020439111524158053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,10240,0.02365511159102122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,8192,0.019506666395399306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,6144,0.017921777235137094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,7168,0.01771111124091678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,3072,128,0.007091555330488417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,7168,0.029902223083708022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,5120,0.015384000208642749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,6144,0.015966221690177917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,4096,0.013089777694808112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,6144,0.026562665899594624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,5120,0.014835556348164877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,3584,0.011648000114493899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,4096,0.012537777423858643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,4096,0.018969777557584975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,3072,0.01036000003417333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,3584,0.011861333416567909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,12288,0.03109333250257704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,2560,0.009020444419648912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,3584,0.017284444636768766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,51200,0.18280000156826445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,3072,0.011239999698268043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,2048,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,2560,0.010022222167915767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,3072,0.01588000026014116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,1536,0.00656355544924736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,2048,0.009385777844323052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,2560,0.013884444203641681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,1024,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,10240,0.03926666577657064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,2048,0.012129777835475074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,768,0.004222222086456087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,8192,0.03311999970012241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,1024,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,512,0.0038373333712418876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,1024,0.008638222184446123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,768,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,256,0.0032186667538351486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,768,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,512,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,128,0.0028293333533737394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,5120,0.02272533377011617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,512,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,256,0.006999111009968652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,64,0.002595555658141772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2560,32,0.0028151110228565005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,256,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,128,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,128,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,65536,0.1694800059000651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,32,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,51200,0.12969422340393066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,65536,0.07658933268653022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,65536,0.18978933493296304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,16384,0.047819554805755615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2560,1536,0.008631111019187504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,51200,0.06088888645172119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,1536,0.010306666294733683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,51200,0.147779557440016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,16384,0.022720000810093347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,10240,0.02715377840730879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,16384,0.049138665199279785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,12288,0.0188746667570538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,8192,0.023016000787417095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,7168,0.02069244450993008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,10240,0.01719377769364251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,10240,0.03310222095913357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,6144,0.01733688844574822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,8192,0.02797244323624505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,7168,0.014090667168299357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,5120,0.01555022266176012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,7168,0.025303999582926433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,6144,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,4096,0.012993777791659037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,6144,0.02234755622016059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,5120,0.011011555790901184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,3584,0.011924444801277585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,5120,0.01924977699915568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,4096,0.009911999934249455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,3072,0.010392000277837118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,4096,0.016136000553766888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,3584,0.01479466590616438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,2560,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,3072,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,12288,0.03410844339264764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,3072,0.01371022231049008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,2048,0.008065777520338694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,2560,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,1536,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,2560,0.012106666962305704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,2048,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,1024,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,8192,0.01407911048995124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,2048,0.010680889089902243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,768,0.0046311111913787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,1536,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,1536,0.009004444711738164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,1024,0.007280888656775157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,512,0.0038702223036024305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,768,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,1024,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,256,0.0032133333798911837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,768,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,512,0.006552888701359431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,128,0.0028231110837724474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,256,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,512,0.006678222368160884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,64,0.0025555555605226094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,256,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,128,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,128,0.006218666831652324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,65536,0.12029511398739284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,64,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,32,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,51200,0.11357066366407607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,65536,0.07311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,16384,0.036100443866517805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,65536,0.18008000320858428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2048,12288,0.038457777765062116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,51200,0.059325330787234835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,51200,0.14203022585974798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,16384,0.022134222918086584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,10240,0.02237422267595927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,16384,0.04755111204253303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,12288,0.018212444252438016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,12288,0.03713066710366143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,8192,0.02032888929049174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,2560,64,0.006490666833188798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,10240,0.016864000095261466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,10240,0.03223022156291538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,8192,0.013471110827393003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,2048,32,0.0028373334142896864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,6144,0.016139555308553908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,8192,0.027056889401541814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,2048,3584,0.009371555513805812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,7168,0.012438221938080259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,5120,0.01369511087735494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,7168,0.024096000525686476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,4096,0.011709333293967776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,6144,0.011423110961914062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,6144,0.020993777447276648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,5120,0.0105591110057301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,3584,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,3072,0.009540444446934594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,4096,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,5120,0.018231110440360177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,3584,0.009139555195967356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,2560,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,4096,0.016099555624855887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,2048,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,3584,0.014644443988800049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,12288,0.02630133264594608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,3072,0.01314311143424776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,1536,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,2048,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,2560,0.011733333269755045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,1024,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,1536,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,2048,0.010456889039940305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,768,0.0042160000238153665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,1536,0.00886933339966668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,1024,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,1024,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,768,0.006999111009968652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,7168,0.01811377704143524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,256,0.0028560000161329904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,768,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,128,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,512,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,256,0.006340444501903322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,512,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,64,0.002518222149875429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,32,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,128,0.006221333311663733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,128,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,65536,0.08754933542675442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,64,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,32,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,51200,0.07870755592981975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,3072,0.009053332938088311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,65536,0.07317599985334608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1536,2560,0.008512000242869059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,65536,0.1392480002509223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,51200,0.05834844377305773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,12288,0.025576000412305195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,51200,0.10909777879714966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,16384,0.021523555119832356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,10240,0.022297778063350256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,16384,0.03795288999875387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,12288,0.01756266587310367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1536,512,0.0035128887328836652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,8192,0.018266666266653273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,12288,0.029911110798517864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,10240,0.01592088904645708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,10240,0.02602488961484697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,8192,0.013112000293201871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,6144,0.012864000267452665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,7168,0.012079111403889127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,8192,0.021672889590263367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,5120,0.011528000235557556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,7168,0.0196115556690428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,6144,0.011351111034552256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,6144,0.017645332548353408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,5120,0.015488889482286243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,3584,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,4096,0.01331999980741077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,16384,0.027320888307359483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,3584,0.00906577789121204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,3072,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,3584,0.012378666963842181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,2560,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,3072,0.008707555631796518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,3072,0.011365333365069495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,2048,0.005519999812046687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,7168,0.01663733356528812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,2560,0.010205333431561789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,1536,0.00509333320789867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,2048,0.008164444731341468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,2048,0.009143999881214565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1536,256,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,1024,0.0038737778862317405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,1536,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,4096,0.009124444590674506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,1536,0.007927999728255803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,768,0.0034719999465677473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,5120,0.01035822182893753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,512,0.002871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,1024,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,768,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,4096,0.009334222310119206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,1024,0.006990222467316522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,768,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,512,0.0063475556671619415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,128,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,512,0.006113777971929974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,256,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,256,0.005636444522274866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,128,0.006230222268237009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,32,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,128,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,65536,0.06742488675647311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,64,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,1024,2560,0.008395555946562026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,1024,32,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,51200,0.05223733186721802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,65536,0.07259377506044176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,16384,0.02201333310869005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,65536,0.13457600275675455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,51200,0.05772888660430908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,12288,0.018220444520314533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,16384,0.021034666233592566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,10240,0.017143999536832173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,51200,0.10495022270414565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,16384,0.036842667394214206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,12288,0.017218665944205392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,8192,0.014349333114094205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,12288,0.029186666011810303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,10240,0.015511999527613321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,7168,0.013066666821638743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,256,0.002537777854336633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,8192,0.013140444126394061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,10240,0.024792888098292883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,6144,0.011389333340856763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,5120,0.009882666998439366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,8192,0.020534222324689228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,7168,0.012062222593360476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,1024,64,0.002238222294383579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,6144,0.011030221978823343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,7168,0.0188044442070855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,3584,0.007245333658324347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,6144,0.016872889465755887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,4096,0.009314667019579146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,5120,0.014855111638704935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,3072,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,4096,0.012710221939616732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,3584,0.00904266701804267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,2560,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,3584,0.011808888779746162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,2048,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,3072,0.011032889286677042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,2560,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,1536,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,2560,0.009947555760542551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,2048,0.00794400026400884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,1024,0.0035662220584021676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,1536,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,1536,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,768,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,1024,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,512,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,1024,0.006873777757088344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,768,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,256,0.0026355555488003623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,4096,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,512,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,128,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,5120,0.01036000003417333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,512,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,64,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,256,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,128,0.006226666685607698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,768,32,0.002193777718477779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,128,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,65536,0.045026666588253446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,64,0.005312888986534542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,51200,0.03714666763941447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,32,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,65536,0.07195378012127347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,16384,0.015448888142903646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,65536,0.09477955765194362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,51200,0.05766044722663032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,12288,0.015487111277050443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,51200,0.07476088735792372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,16384,0.020417778028382193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,16384,0.027059555053710938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,10240,0.014125333891974555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,12288,0.017092444830470614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,8192,0.015050666199790107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,10240,0.015767999821239047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,768,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,7168,0.012118221984969245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,10240,0.018681777848137748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,8192,0.013015111287434896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,256,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,6144,0.010088000032636855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,8192,0.016188444362746347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,7168,0.011794666449228922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,5120,0.00872622181971868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,7168,0.014826666977670459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,6144,0.011019555230935415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,768,3072,0.008658666577604082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,4096,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,6144,0.013482666677898832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,3584,0.0068853331936730286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,5120,0.010212444596820408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,5120,0.012016889121797351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,3072,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,4096,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,3584,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,2560,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,3584,0.010015111002657149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,3072,0.008644444247086843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,2048,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,3072,0.009338666995366415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,2560,0.008067555725574493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,1536,0.0041884444653987885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,2560,0.008738666772842407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,2048,0.007724444071451823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,12288,0.02181333303451538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,1024,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,2048,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,1536,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,768,0.003157333367400699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,1536,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,1024,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,512,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,768,0.00664533343580034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,1024,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,256,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,768,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,512,0.006246222390068903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,128,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,512,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,64,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,128,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,256,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,4096,0.009309333231714036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,512,32,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,128,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,65536,0.02968177861637539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,768,2048,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,51200,0.027492443720499676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,32,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,16384,0.0109120003051228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,65536,0.09334666861428155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,51200,0.055640889538658984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,51200,0.07320533196131389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,12288,0.011011555790901184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,16384,0.020113777783181932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,10240,0.010219555762079027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,16384,0.026216889421145122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,12288,0.016934222645229764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,8192,0.009076444639099969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,12288,0.02086666723092397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,10240,0.015511999527613321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,7168,0.008719999757077958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,10240,0.018618666463428073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,8192,0.01272266689274046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,6144,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,8192,0.015787555111779105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,512,256,0.006228444476922353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,5120,0.008451555338170793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,7168,0.014488889111412896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,6144,0.010951999988820819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,4096,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,5120,0.010189333723651038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,6144,0.013083555632167392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,3584,0.006545777950021956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,5120,0.011749332977665795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,512,64,0.004963555683692296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,3072,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,65536,0.0715893308321635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,3584,0.008982222113344405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,2560,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,3072,0.008625778059164682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,3584,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,2048,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,3072,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,2560,0.00816000004609426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,1536,0.0038782221575578055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,2560,0.008350222474998897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,2048,0.00776622196038564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,1024,0.0031964443624019623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,2048,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,1536,0.007357333269384172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,768,0.0028719999310043124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,1536,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,1024,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,512,0.002530666689078013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,768,0.006698666761318843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,1024,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,256,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,768,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,128,0.0021262222694026097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,7168,0.011761777930789523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,512,0.005389333185222413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,256,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,64,0.0018604445374674266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,256,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,128,0.006309333360857434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,256,32,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,128,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,4096,0.00925511121749878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,4096,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,32,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,65536,0.07031555308236016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,65536,0.09257600042555068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,16384,0.009304888546466827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,51200,0.05458577805095249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,51200,0.07362044519848294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,12288,0.00815644446346495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,16384,0.01991555591424306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,16384,0.025770665870772466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,10240,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,12288,0.016574222180578444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,256,512,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,12288,0.02088266611099243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,8192,0.009373333719041612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,10240,0.015290665957662793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,7168,0.008746667040718926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,10240,0.01819022165404426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,8192,0.012741333080662621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,6144,0.008001777860853408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,65536,0.024013333850436743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,8192,0.015837333268589444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,7168,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,5120,0.007370666497283512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,256,64,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,7168,0.014259555273585849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,6144,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,4096,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,3584,0.006230222268237009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,5120,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,6144,0.01312711089849472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,3072,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,4096,0.009304888546466827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,3584,0.009003555609120263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,4096,0.010168888502650792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,2560,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,3072,0.008520888785521189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,3584,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,2048,0.004563555535342959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,3072,0.008963555925422246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,1536,0.0038248888320393036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,2560,0.00830488900343577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,2048,0.007711999946170383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,1024,0.0032008888406885993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,2048,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,1536,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,768,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,1536,0.006755555669466655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,512,0.0024871110088295406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,1024,0.006909333169460297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,768,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,256,0.0021964444054497615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,512,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,512,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,128,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,256,0.006040888941950268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,256,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,64,0.0019164443429973389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,5120,0.011694221860832639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,128,0.006221333311663733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,32,0.0021093332519133887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,128,0.004987555659479565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,64,0.004942222187916438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,32,0.00508355556262864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,51200,0.017471111483044095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,16384,0.00977955593003167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,65536,0.09202311436335246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,12288,0.007954667011896769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,51200,0.07308977842330933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,10240,0.00701155596309238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,16384,0.02568977740075853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,8192,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,1024,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,12288,0.020639111598332722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,7168,0.00665244460105896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,128,768,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,6144,0.0063031112982167145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,5120,0.008066666622956594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,10240,0.018182223041852314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,4096,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,8192,0.015474667151769003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,3584,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,6144,0.013070222404268054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,3072,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,5120,0.01146044416560067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,65536,0.022209778428077698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,2560,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,4096,0.010096888575288985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,2048,0.004211555752489301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,1536,0.0036106668412685394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,3072,0.008930666579140557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,769,128,2560,0.00830488900343577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,1024,0.0031519999934567344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,2560,0.008212444682916006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,2048,0.007583111524581909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,768,0.002838222309947014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,1536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,512,0.0024622221373849446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,256,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,1024,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,128,0.0019404445257451797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,768,0.005877333382765452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,64,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,512,0.005436444448100195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,256,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,128,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,65536,0.019263111882739596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,64,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,51200,0.012668444050682915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,32,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,16384,0.01517599986659156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,65536,0.0925146672460768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,3584,0.009892444643709395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,51200,0.07277422481113009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,12288,0.01275288893116845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,10240,0.008007999923494127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,16384,0.02551288902759552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,8192,0.010145777629481422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,7168,0.01752711170249515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,12288,0.02064622276359134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,6144,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,8192,0.015483554866578845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,5120,0.013384000294738345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,7168,0.01423111061255137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,4096,0.0075866662793689305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,64,32,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,6144,0.013080000049538083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,3584,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,3072,0.006082666830884085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,5120,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,2560,0.005535111245181825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,4096,0.01015022231472863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,64,7168,0.014148443937301636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,2048,0.004859555512666702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,3584,0.009823111196359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,3072,0.008907555705971187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,1024,0.0032239999208185407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,2560,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,768,0.0028008888992998335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,2048,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,512,0.002475555572244856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,1536,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,256,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,1024,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,128,0.0018915554715527429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,768,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,512,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,10240,0.018170666363504197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,256,0.005377777748637729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,128,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,64,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,769,32,32,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,128,51200,0.017992888887723286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,769,32,1536,0.00407644444041782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,16384,0.8911128573947482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,16384,0.32866933610704213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,16384,1.0151404274834526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,12288,0.705984910329183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,12288,0.740296893649631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,51200,1.0830329259236653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,12288,0.24633333418104383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,51200,2.953810585869683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,10240,0.5558151139153374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,16384,0.8600168757968478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,8192,0.4773884349399143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,10240,0.6846524344550239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,12288,0.6283155547247993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,51200,3.4284960428873696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,10240,0.21103555626339385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,8192,0.17223821745978463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,8192,0.5274044672648112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,7168,0.39790932337443036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,7168,0.42183645566304523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,6144,0.3259893258412679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,7168,0.1555066638522678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,10240,0.5346586439344618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,6144,0.4023439884185791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,5120,0.34462844000922305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,6144,0.13346666759914821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,7168,0.3832968870798747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,5120,0.32415111859639484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,8192,0.43666577339172363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,5120,0.12305333879258896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,4096,0.2363591194152832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,6144,0.34069866604275173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,3584,0.23647377226087782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,4096,0.0959706637594435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,4096,0.2484346760643853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,5120,0.2727484438154432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,3072,0.17329866356319854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,3584,0.2255049016740587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,4096,0.22762489318847656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,3072,0.07556711302863227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,3072,0.18924000528123644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,2560,0.1396311124165853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,3072,0.17365777492523193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,2560,0.06595910920037164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,2048,0.11641777886284722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,2560,0.16136622428894043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,2560,0.1502337720659044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,1536,0.09956000248591106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,2048,0.05664622121387058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,2048,0.13237867090437147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,2048,0.12575199868943956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,3584,0.09921955400043064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,1536,0.10429333315955268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,1536,0.10036355257034302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,3584,0.20657422807481554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,768,0.04970222380426195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,1024,0.07943644126256307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,1024,0.03643200132581923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,768,0.0328435566690233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,768,0.06435377730263604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,512,0.039685332112842135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,768,0.0659431086646186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,512,0.030049777693218652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,512,0.05590133534537422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,256,0.03018933203485277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,512,0.05450310972001818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,128,0.02279022170437707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,256,0.027887112564510767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,256,0.05076800121201409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,256,0.04350489046838549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,128,0.027670222851965163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,64,0.02038666605949402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,1024,0.07955644528071086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,32,0.019588443968031142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,128,0.03395288851526048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,64,0.04677599999639723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,32,0.0466177761554718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,65536,1024,0.06779555479685466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,65536,128,0.0473911099963718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,51200,2.991585837470161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,65536,1.0834862391153972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,65536,3.2684728834364147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,51200,2.48709593878852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,1536,0.04468533396720886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,65536,3.329844368828667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,51200,2.657668431599935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,16384,0.26158044073316783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,16384,0.7497537930806478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,12288,0.5305040147569444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,16384,0.7889306810167102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,12288,0.5972586737738715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,12288,0.1989048851860894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,10240,0.4205306635962592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,16384,0.672393798828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,10240,0.49756977293226456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,10240,0.16313955518934461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,8192,0.35229333241780597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,51200,0.8221244282192655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,12288,0.5228613217671713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,8192,0.3819288942548964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,7168,0.3815235561794705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,10240,0.4313093291388617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,8192,0.13470044400956896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,7168,0.35498489273919004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,8192,0.33896711137559676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,6144,0.27297867668999565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,7168,0.1225831111272176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,6144,0.2937946584489611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,5120,0.2221004433102078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,6144,0.10552177826563518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,5120,0.24807198842366537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,7168,0.305984894434611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,4096,0.22032088703579375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,6144,0.27147822909884983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,3584,0.157096889283922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,5120,0.09168799718221028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,4096,0.08059110906389025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,5120,0.22203555372026232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,4096,0.1972479952706231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,4096,0.18220888243781197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,3072,0.133206221792433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,3584,0.068631112575531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,3584,0.17447377575768364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,2560,0.14018577999538845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,3072,0.1467004484600491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,3584,0.1620604462093777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,3072,0.05980000231001112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,2048,0.10933422380023533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,65536,3.031040827433268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,3072,0.1412808895111084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,2560,0.052962667412228055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,2560,0.12678578164842394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,51200,2.429040061102973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,1536,0.08544266886181301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,2048,0.10498399866951837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,2560,0.12193689081403945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,1024,0.05412266651789347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,2048,0.04821333289146423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,2048,0.10097777843475342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,1536,0.043544001049465604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,1024,0.02993511160214742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,768,0.04549511273701986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,1024,0.06396889024310641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,1024,0.06542666753133138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,1536,0.08289421929253472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,768,0.027541332774692114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,768,0.056959999932183154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,768,0.054767999384138316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,512,0.025103999508751765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,512,0.04498044318623013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,512,0.045831110742357045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,128,0.018352000249756705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,256,0.02331200076474084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,256,0.036883556180530123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,128,0.03804888990190294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,64,0.016012445092201233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,51200,128,0.022724444667498272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,32,0.01703733371363746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,1536,0.08341688579983181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,64,0.03764355513784621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,512,0.029616889026429918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,32,0.03763466742303636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,51200,256,0.023073777556419373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,51200,256,0.040956444210476346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,65536,0.9783973693847656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,65536,0.36475290192498105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,51200,128,0.028220444917678833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,51200,0.7621075312296549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,65536,1.046778678894043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,16384,0.2733084360758464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,51200,0.28444888856675893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,16384,0.09364444679684109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,12288,0.1870408852895101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,16384,0.286352899339464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,51200,0.8425146738688151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,10240,0.17313244607713488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,12288,0.0723279979493883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,16384,0.24375467830234102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,12288,0.19620977507697213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,8192,0.12668532795376247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,10240,0.1575982173283895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,12288,0.18627377351125082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,10240,0.059947556919521756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,7168,0.12654577361212835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,10240,0.1597999996609158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,8192,0.049624890089035034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,8192,0.1300391091240777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,6144,0.11653867032792832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,7168,0.11491821871863471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,7168,0.04504444532924228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,5120,0.0989004439777798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,65536,0.9864675733778211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,6144,0.039763556586371526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,7168,0.1142115592956543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,4096,0.07788978020350139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,6144,0.10014666451348199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,5120,0.034416000048319496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,51200,0.7667457792494031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,3584,0.06469155682457818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,4096,0.02935911218325297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,4096,0.06803555621041192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,5120,0.0832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,5120,0.08816266722149319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,3072,0.05832799937989977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,3584,0.026165333059098985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,3584,0.06108178032769097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,2560,0.047713776429494224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,3584,0.06604888704088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,3072,0.02368800010946062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,3072,0.05458844370312161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,8192,0.12792710463205972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,3072,0.058261334896087646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,2560,0.021592888567182753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,2560,0.046471112304263644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,6144,0.09761866596009995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,1536,0.029472000069088403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,2560,0.05162133442031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,2048,0.01874755488501655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,2048,0.0381279985109965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,1024,0.02056799994574653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,2048,0.044271111488342285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,1536,0.030709332889980737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,768,0.016333333320087857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,1536,0.03700977894994948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,1024,0.013770666387346057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,1024,0.024177778098318312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,1024,0.030251555972629126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,4096,0.07264177666770087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,768,0.012480888929631976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,768,0.020218667056825426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,256,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,768,0.02568000058333079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,512,0.011418666276666852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,512,0.016925333274735343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,128,0.009488888912730748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,512,0.02172800070709652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,256,0.01070844464831882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,256,0.015152888165579902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,2048,0.03568799959288703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,64,0.008958222137557136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,256,0.019159111711714003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,32,0.008973333570692275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,128,0.014499555031458536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,128,0.015203555425008139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,64,0.014649778604507446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,1536,0.016501333978441026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,16384,32,0.014781332678265043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,16384,512,0.012415111064910889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,65536,0.6702115800645617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,51200,0.49833064609103733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,65536,0.2570462226867676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,65536,0.8035635418362088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,16384,0.18785333633422852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,128,0.010761777559916178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,51200,0.6379929118686253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,16384,0.0689493351512485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,16384,0.19354489114549425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,12288,0.1170204480489095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,16384,0.18647289276123047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,10240,0.13080710834927028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,12288,0.053242666853798755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,12288,0.13376711474524602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,65536,0.6853617562188042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,10240,0.04516266783078512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,10240,0.12184088759952122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,8192,0.0892595582538181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,10240,0.11390577422247992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,7168,0.08117955260806613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,8192,0.03666755557060242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,8192,0.09797510835859512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,51200,0.20657600296868217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,6144,0.06488711304134794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,8192,0.09416444434059991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,7168,0.033238222201665245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,7168,0.08795999818378025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,5120,0.06531555785073175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,7168,0.08675822284486558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,6144,0.029480000336964924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,6144,0.07672622468736437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,4096,0.05381866627269321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,5120,0.06496089034610324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,5120,0.026014222039116755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,3584,0.04599822229809231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,5120,0.06611377663082547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,4096,0.0221760008070204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,4096,0.05496266815397474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,3072,0.03924177752600776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,3584,0.019987555013762582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,3584,0.048398223188188344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,51200,0.5515519777933756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,2560,0.028872887293497723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,3584,0.050294223758909434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,3072,0.01852888862291972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,3072,0.04222044348716736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,2048,0.026242666774325903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,3072,0.044694221682018705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,2560,0.036173333724339805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,2560,0.017115554875797696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,2560,0.03989244500796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,1536,0.020163555939992268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,2048,0.015134221977657743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,2048,0.030662222041024104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,1024,0.014222222897741528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,2048,0.03369066781467862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,1536,0.013295999831623502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,12288,0.14518845081329346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,768,0.011805333197116852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,1536,0.029554665088653564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,1024,0.011025778121418424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,4096,0.05415466758939955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,1024,0.01990133358372582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,512,0.00923555592695872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,1024,0.023191110955344305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,768,0.01015733347998725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,768,0.01668355531162686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,256,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,768,0.020994666549894545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,512,0.013659555878904132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,512,0.009638222555319468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,128,0.006194666855865055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,512,0.017994667092959087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,256,0.00907466643386417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,64,0.005450666778617435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,256,0.01257955531279246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,12288,32,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,256,0.01614755557643043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,128,0.009191999832789103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,128,0.012311999996503195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,128,0.013398222625255585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,64,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,32,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,6144,0.07438221904966566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,65536,0.5735244221157497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,51200,0.4236702124277751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,65536,0.24266934394836426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,16384,0.13928977648417154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,65536,0.6845244301689996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,51200,0.537576887342665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,16384,0.16383288966284856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,16384,0.06259200308057997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,12288,0.10836622450086807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,16384,0.16850222481621635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,12288,0.12288533316718207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,10240,0.08932089143329197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,12288,1536,0.025037333369255066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,12288,0.04872177706824409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,12288,0.12991555531819662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,65536,0.662759992811415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,8192,0.07580533292558458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,10240,0.042248000701268516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,10240,0.10345066918267144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,10240,0.11072799894544814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,8192,0.034640000926123724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,8192,0.08870133426454331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,8192,0.09221688906351726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,51200,0.19364710648854574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,7168,0.03229511115286086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,6144,0.0667475528187222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,7168,0.075254221757253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,5120,0.05308622121810913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,6144,0.02782400117980109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,6144,0.06509600083033244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,6144,0.07262844509548612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,5120,0.02420355545149909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,4096,0.042245331737730235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,5120,0.05577422512902153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,5120,0.06309866905212402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,51200,0.5150675773620605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,3584,0.04351644383536445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,4096,0.04606489009327359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,4096,0.052896890375349254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,7168,0.07741955253813002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,3584,0.040673779116736516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,3072,0.03454577922821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,3584,0.01939911146958669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,3584,0.04801511102252536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,2560,0.031152000029881794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,3072,0.017879999346203275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,2048,0.024309333827760484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,2560,0.016566221912701923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,3072,0.043248001072141856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,2560,0.031409776873058744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,1536,0.019704000817404855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,2560,0.03798933492766486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,2048,0.014665777484575907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,2048,0.026612444056404963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,1024,0.0147733340660731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,2048,0.03294310967127482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,1536,0.02180533276663886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,1536,0.01276266657643848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,1536,0.02833066715134515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,768,0.012044444680213928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,4096,0.021173333128293354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,1024,0.010716444088353051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,1024,0.017251556118329365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,512,0.008997333546479544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,1024,0.023149333066410486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,768,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,768,0.014728888869285583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,256,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,512,0.009504000345865885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,768,0.019903999235894945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,512,0.012260444462299347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,128,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,512,0.0173688895172543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,256,0.008721777962313758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,256,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,3072,0.03616177704599168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,256,0.01479644411140018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,128,0.009017777939637503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,128,0.011010666688283285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,7168,0.08329777585135566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,128,0.012010667059156628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,64,0.011029333704047732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,10240,32,0.011043555206722684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,65536,0.4351377752092149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,65536,0.1928346686893039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,51200,0.38808533880445695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,65536,0.5494497617085775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,16384,0.13779200447930232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,51200,0.42321867412990993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,51200,0.15216978391011557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,12288,0.10198577907350327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,16384,0.05053066545062595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,64,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,10240,32,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,16384,0.1417048904630873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,10240,0.09288089142905341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,12288,0.04097422295146518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,65536,0.5148186683654785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,12288,0.10254489050971137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,8192,0.0650764438841078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,10240,0.03439377744992574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,12288,0.10565777619679768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,51200,0.4054764376746283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,10240,0.08649866448508368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,7168,0.0618399977684021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,10240,0.09096978108088176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,8192,0.02790400054719713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,8192,0.07161688804626465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,6144,0.05774755610360039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,7168,0.06363644202550252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,8192,0.08306755622227986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,7168,0.06708888875113593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,5120,0.04411377840571933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,6144,0.022859555151727464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,6144,0.05925955375035604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,5120,0.02048177851570977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,16384,0.13510578208499485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,5120,0.04690577917628818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,5120,0.05260355605019463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,3584,0.03084355592727661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,4096,0.03932799895604452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,4096,0.044066667556762695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,3584,0.03512800070974562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,3072,0.026367111338509455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,3584,0.016694222887357075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,3584,0.040008889304267034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,7168,0.025546666648652818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,2560,0.023287110858493384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,3072,0.015522667103343539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,3072,0.03115733464558919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,3072,0.035976890060636736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,6144,0.05472977624999153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,2048,0.019527999891175162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,2560,0.014194667339324951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,2560,0.02749155627356635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,1536,0.014197332991494073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,4096,0.0370488895310296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,2560,0.03200533323817783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,2048,0.013023110727469126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,1024,0.012244444754388599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,2048,0.023059555225902136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,4096,0.01812533371978336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,1536,0.01108266661564509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,1536,0.019147555033365887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,1536,0.024367999699380662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,1024,0.009528888596428765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,1024,0.014829332629839579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,1024,0.01960711181163788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,768,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,768,0.017665778597195942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,512,0.011016888750924004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,256,0.006568000134494569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,512,0.01550222271018558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,256,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,256,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,128,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,256,0.014103111293580798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,2048,0.02804888950453864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,128,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,64,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,32,0.005526222288608551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,512,0.00815822184085846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,64,0.009725333087974125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,768,0.012961777547995249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,8192,32,0.009698666632175446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,512,0.00868888861603207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,65536,0.4342319965362549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,65536,0.18403022819095186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,65536,0.4935262468126085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,51200,0.3820942242940267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,128,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,8192,768,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,16384,0.12623111406962076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,128,0.011254222028785281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,51200,0.3908577760060628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,51200,0.1435155603620741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,12288,0.09631733099619548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,16384,0.053499556250042386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,16384,0.12066045072343613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,65536,0.5024782286749946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,12288,0.03850311040878296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,12288,0.09261866410573323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,16384,0.1338497797648112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,51200,0.3971262243058946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,12288,0.10296089119381374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,10240,0.03320711188846164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,10240,0.07883199718263414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,7168,0.07150933477613661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,10240,0.0897573365105523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,8192,0.026985777748955622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,8192,0.06466488705741034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,6144,0.056461334228515625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,7168,0.02481422159406874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,7168,0.05791999896367391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,7168,0.0658880008591546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,5120,0.05032533407211304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,6144,0.05024622215165032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,6144,0.022537777821222942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,4096,0.03776622149679396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,5120,0.01993688941001892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,5120,0.04316888915167915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,5120,0.05152444375885857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,10240,0.0982888870769077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,4096,0.036152889331181846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,4096,0.017633777525689866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,4096,0.04351644383536445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,8192,0.07326222128338285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,3584,0.016352888610627916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,3072,0.02885422110557556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,3584,0.032384001546435885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,3584,0.03991377684805129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,2560,0.02458488941192627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,3072,0.02897777822282579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,3072,0.03564800156487359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,2560,0.025547555751270715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,2560,0.013810666898886362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,2048,0.020603555772039626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,2560,0.03131999903255039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,2048,0.012333333492279053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,2048,0.02167822255028619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,1536,0.016148444679048326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,2048,0.02773244513405694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,3584,0.03376355436113145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,1024,0.01035288886891471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,1536,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,1536,0.017729777428838942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,1536,0.02400711178779602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,768,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,1024,0.013600889179441663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,1024,0.009315555294354757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,3072,0.015090665883488126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,512,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,1024,0.01928711103068458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,768,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,768,0.011995555626021491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,256,0.005146666533417172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,768,0.01767200065983666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,512,0.010403555300500657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,6144,0.058054222000969775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,128,0.004691555682155821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,512,0.014977777997652689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,256,0.009588444398509132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,256,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,64,0.004208000169859992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,256,0.013691555294725629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,7168,32,0.004243555582231945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,128,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,128,0.009085333181752099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,128,0.011154666543006897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,64,0.009316444396972656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,7168,32,0.0090524446633127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,65536,0.3830408785078261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,51200,0.2593279944525825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,8192,0.07310489151212904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,65536,0.1431244479285346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,65536,0.4234853320651584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,16384,0.09285600317849053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,512,0.008350222474998897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,51200,0.11191466119554307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,12288,0.06769777668846978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,16384,0.03840444485346476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,16384,0.10522400008307563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,10240,0.06272266970740424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,16384,0.1007688906457689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,12288,0.029888888200124104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,65536,0.3712880081600613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,8192,0.04943644338183933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,51200,0.3018053372701009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,12288,0.0808124476008945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,10240,0.025946666797002155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,10240,0.06788088877995808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,7168,0.045175111956066556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,8192,0.021698666943444148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,8192,0.056223114331563316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,6144,0.03719022207789951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,8192,0.058435552650027804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,7168,0.04954222175810072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,7168,0.019900444481107924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,5120,0.029674665795432195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,7168,0.051607112089792885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,6144,0.017972444494565327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,6144,0.04321066538492838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,51200,0.3338079982333713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,4096,0.02566755480236477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,6144,0.046911110480626426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,5120,0.016192888220151264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,5120,0.037224001354641385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,3584,0.024124445186720953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,5120,0.03966755668322245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,4096,0.014525334040323893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,4096,0.031207998593648274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,3072,0.020908445119857788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,3584,0.013495999905798169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,4096,0.0354017780886756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,3584,0.02819288770357768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,12288,0.08340089188681708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,2560,0.01688799924320645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,3072,0.012470222181744046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,2048,0.014860444598727755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,3072,0.025145777397685583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,2560,0.011692444483439127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,3072,0.02849422229660882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,2560,0.022133333815468684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,10240,0.0686195558971829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,2560,0.0258568889564938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,2048,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,2048,0.019023999571800232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,1024,0.009297777381208207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,2048,0.02318488889270359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,1536,0.01551466683546702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,1536,0.019488000207477145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,768,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,1024,0.008329778081840938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,1024,0.01626666718059116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,768,0.011016888750924004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,768,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,512,0.006008000009589725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,768,0.015127999915017022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,3584,0.031806220610936485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,512,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,256,0.004604444321658877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,512,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,512,0.013784888717863293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,128,0.0038782221575578055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,256,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,256,0.008984888593355814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,256,0.012089778151777057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,1536,0.009218666288587783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,128,0.008678221868144141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,32,0.0038515557017591265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,128,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,1024,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,64,0.008658666577604082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,6144,32,0.008652444514963362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,65536,0.32582399580213756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,51200,0.22155022621154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,65536,0.3708968957265218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,1536,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,6144,64,0.003547555456558863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,65536,0.363306654824151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,51200,0.10673244131935967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,16384,0.09272266758812799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,12288,0.0697751111454434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,16384,0.03680800067053901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,16384,0.09113688601387872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,16384,0.09942133559121026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,12288,0.07054933574464586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,12288,0.029551111989551123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,12288,0.07752622498406304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,65536,0.13275021976894802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,8192,0.045928888850741915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,10240,0.025593777497609455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,10240,0.06010933054818047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,10240,0.06792710887061225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,7168,0.0392577780617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,8192,0.02104444470670488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,8192,0.04888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,128,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,8192,0.05617600017123752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,6144,0.03199111090766059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,7168,0.019263111882739596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,7168,0.044016887744267784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,5120,0.028616888655556574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,7168,0.05077155431111654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,6144,0.017588444881969027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,51200,0.28500088055928546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,4096,0.02418311105834113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,5120,0.015933333171738517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,5120,0.033357332150141396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,10240,0.04767555660671658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,3584,0.020089776979552377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,5120,0.03940977652867635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,4096,0.01423911088042789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,4096,0.0281333327293396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,3072,0.01903555625014835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,4096,0.03364355696572198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,3584,0.013385777672131857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,3584,0.025768889321221247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,3584,0.03143644332885742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,2560,0.016572443975342643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,3072,0.022943110929595098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,3072,0.02812444501452976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,2560,0.011131555669837527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,2560,0.020032000210550096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,2560,0.025080889463424683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,6144,0.038380444049835205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,2048,0.010032888915803697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,2048,0.017274666163656447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,1536,0.01167555567291048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,2048,0.022689777943823073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,6144,0.04559288753403557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,1024,0.009047110875447592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,1536,0.00902844468752543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,1536,0.01420177850458357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,768,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,1536,0.01946311030122969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,1024,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,1024,0.011362666885058085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,512,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,1024,0.016389333539538913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,768,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,768,0.010207110808955299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,256,0.005187555733654234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,768,0.015000889698664347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,512,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,512,0.008804444637563493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,51200,0.29289600584242076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,512,0.013740444348918067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,128,0.004253333227501975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,256,0.007160888777838812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,256,0.008466666771305932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,64,0.0038720000949170855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,256,0.012087110843923358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,128,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,32,0.004512000001139111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,128,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,128,0.00978488889005449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,64,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,5120,32,0.007848888635635376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,65536,0.2533128791385227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,51200,0.24196532037523058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,65536,0.31403644879659015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,65536,0.1228062179353502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,16384,0.07582577731874254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,51200,0.25511733690897626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,3072,0.012248000337017907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,51200,0.09797155857086182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,12288,0.05738933218850029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,65536,0.36207111676534015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,5120,2048,0.014094221923086377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,16384,0.07979644669426812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,10240,0.04740800129042732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,12288,0.027671110298898485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,51200,0.2841946548885769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,16384,0.1001253326733907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,12288,0.061647110515170626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,8192,0.037047111325793795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,12288,0.07755200068155925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,10240,0.024744888146718342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,10240,0.052468445565965444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,7168,0.035063998566733465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,8192,0.020795555578337777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,10240,0.06611644559436373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,6144,0.02717155549261305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,8192,0.055665777789221875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,7168,0.018989332848125033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,7168,0.0388648874229855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,5120,0.024375999967257183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,7168,0.05015288790067037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,6144,0.017324444320466783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,6144,0.033871110942628645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,4096,0.019711999429596793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,6144,0.0442808899614546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,5120,0.01552977826860216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,5120,0.02985422147644891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,3584,0.017873777283562552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,5120,0.039071109559800886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,16384,0.03551200032234192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,4096,0.01352355546421475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,4096,0.025224000215530396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,3072,0.016551110479566786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,4096,0.03329688972897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,3584,0.012799999780125089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,2560,0.013783111340469785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,3584,0.02351644469632043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,3072,0.011695110963450538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,3584,0.03070844544304742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,2048,0.012372444073359171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,3072,0.020639111598332722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,2560,0.010609777437316047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,3072,0.027489778068330552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,2560,0.01774666706720988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,8192,0.04301333427429199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,2048,0.009668444593747457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,2560,0.024541333317756653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,1024,0.006994666324721442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,2048,0.015508444772826301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,2048,0.022273777259720698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,1536,0.009003555609120263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,768,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,1024,0.008070222205585903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,1536,0.01867555578549703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,1024,0.010392000277837118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,1024,0.016137777103318107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,768,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,768,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,256,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,768,0.014485332700941296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,512,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,512,0.01346933344999949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,256,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,256,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,256,0.012049777640236748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,128,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,128,0.007337777978844113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,64,0.003411555455790626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,1536,0.009332444104883406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,128,0.00982844415638182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,1536,0.012759110993809171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,64,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,32,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,65536,0.26505690150790745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,512,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,51200,0.1714551051457723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,65536,0.12160355514950222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,4096,512,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,65536,0.27363377147250706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,16384,0.05732444259855482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,51200,0.21330932776133218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,128,0.0035182221068276297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,51200,0.09723110993703206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,12288,0.05082577798101637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,65536,0.3600462277730306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,16384,0.034513778156704374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,16384,0.07027910815344916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,51200,0.29860178629557294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,4096,32,0.0034942221310403612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,10240,0.03592355714903937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,16384,0.09907288683785333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,12288,0.02753688891728719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,12288,0.0542719993326399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,8192,0.03277955452601115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,12288,0.07618577612770928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,10240,0.04714044597413805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,8192,0.02032355633046892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,8192,0.03869333200984531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,10240,0.06549866994222005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,6144,0.025627555118666753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,8192,0.0546248886320326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,7168,0.018579555882347953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,7168,0.034458667039871216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,5120,0.022063111265500385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,7168,0.0492888887723287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,6144,0.016751999656359356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,6144,0.030626667870415583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,4096,0.019945777124828763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,6144,0.043828444348441235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,5120,0.0151555554734336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,5120,0.03864622116088867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,3584,0.014915555715560913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,4096,0.022670222653283015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,4096,0.032935112714767456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,10240,0.024381332927280005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,3584,0.02074399921629164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,3072,0.013216888739003075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,7168,0.029659556017981634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,3584,0.03019733230272929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,2560,0.011687111523416309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,3072,0.01128088848458396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,3072,0.018588445252842374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,3072,0.027357333236270483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,2560,0.016075554821226332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,2560,0.010430222584141625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,2560,0.023665777511066858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,1536,0.008547555241319869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,5120,0.026537777649031744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,2048,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,2048,0.021339555581410725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,1024,0.006545777950021956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,4096,0.01312800000111262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,1536,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,1536,0.011372444530328115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,768,0.005336888962321811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,1536,0.0185244447655148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,1024,0.008034666379292807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,3584,0.01254755589697096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,1024,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,512,0.004569777597983678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,1024,0.01573777695496877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,768,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,768,0.008638222184446123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,256,0.0036391110883818734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,768,0.014480888843536377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,512,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,128,0.0031928889867332247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,512,0.01313066648112403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,2048,0.009988444546858469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,256,0.011807999677128263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,64,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,128,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,128,0.007339555356237624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3584,32,0.0031439999325407874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,2048,0.01388088862101237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,128,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,65536,0.16907555527157256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,64,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,32,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,51200,0.1740595499674479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,65536,0.08226755592558119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,65536,0.23828445540534127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,16384,0.056468447049458824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,65536,0.22590398788452148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,51200,0.18573689460754395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,51200,0.06554489003287421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,512,0.007335111498832703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,12288,0.04358844293488396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,16384,0.023391111029518977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,51200,0.181532449192471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,16384,0.061069336202409535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3584,256,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,10240,0.035529779063330755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,16384,0.06461511055628459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,12288,0.01924000018172794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,12288,0.0468444459968143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,8192,0.02754755483733283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,10240,0.017669333351982962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,12288,0.05109688970777723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,10240,0.04000711109903123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,8192,0.014803555276658801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,10240,0.044574220975240074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,8192,0.033583111233181424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,8192,0.03825777769088745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,7168,0.013558222187889947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,7168,0.03011911114056905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,7168,0.034448888566758894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,5120,0.018427555759747822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,6144,0.012409778104888068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,6144,0.026610665851169165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,6144,0.03124533428086175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,4096,0.014911999305089315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,5120,0.011210666762457954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,5120,0.023346667488416035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,5120,0.02793333265516493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,256,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,3584,0.013781333135233985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,4096,0.01018488903840383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,4096,0.019899555378490023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,3072,0.012281777958075205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,4096,0.024057777391539678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,3584,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,7168,0.024872000018755596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,2560,0.010942222343550788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,3584,0.022318222456508215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,3072,0.015875554747051664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,3072,0.009072889056470659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,2048,0.009659555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,6144,0.021524444222450256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,2560,0.008667555948098501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,3072,0.019729778170585632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,2560,0.014131555954615274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,1536,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,2560,0.018094221750895183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,2048,0.008447111480765873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,2048,0.012440000143316058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,1024,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,2048,0.016208888755904306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,1536,0.007692444655630324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,1536,0.010344888601038191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,768,0.004616888860861461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,1536,0.014549333188268872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,1024,0.007347555624114142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,1024,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,512,0.0038728887836138406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,1024,0.013066666821638743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,768,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,768,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,256,0.0032586666444937387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,768,0.012708444562223224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,512,0.006827555596828461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,128,0.0028568889117903183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,512,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,512,0.011445333560307821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,3584,0.018458666072951425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,256,0.006593777901596493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,256,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,64,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,3072,32,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,256,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,128,0.006433777925040986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,128,0.006665777828958299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,128,0.00871288859181934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,65536,0.1588080061806573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,64,0.006846222198671765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,3072,32,0.006816000160243776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,51200,0.12518221802181667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,65536,0.07828444242477417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,51200,0.06223377916547987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,65536,0.2239768770005968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,12288,0.037323554356892906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,51200,0.17486488819122314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,16384,0.02289955483542548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,16384,0.058654222223493785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,51200,0.17332622740003797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,10240,0.029880000485314265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,16384,0.06364799870385064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,12288,0.01900800069173177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,12288,0.04496888981925117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,8192,0.02567911148071289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,10240,0.017282666431532968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,12288,0.05050666795836555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,10240,0.03907910982767741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,7168,0.021292444732454088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,10240,0.044066667556762695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,8192,0.03251111176278856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,6144,0.019093333019150626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,8192,0.03720088799794515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,7168,0.013101333545313941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,7168,0.033955554167429604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,5120,0.01590666671593984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,6144,0.012085333466529846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,16384,0.04896266592873467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,6144,0.026138666603300307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,4096,0.01368977791733212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,6144,0.030431999100579157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,5120,0.010882666541470421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,5120,0.022618666291236877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,65536,0.21650311681959364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,3584,0.012744888663291931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,4096,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,4096,0.018910222583346896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,4096,0.023175110419591267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,3072,0.011180444724029966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,3584,0.00942844483587477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,3584,0.017056889004177518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,3584,0.02215911116864946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,2560,0.009927111367384592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,8192,0.014298665854665967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,3072,0.00904177791542477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,3072,0.01594666639963786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,2048,0.008586666650242275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,3072,0.019662222928471036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,7168,0.0295413335164388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,2560,0.008655110994974772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,1536,0.006319111006127463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,2560,0.014106666048367819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,2048,0.008021333151393467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,2048,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,1024,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,2048,0.016189333465364244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,1536,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,1536,0.010079999764760336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,768,0.004407111141416761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,1536,0.01443555619981554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,1024,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,5120,0.026733333865801494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,512,0.0038764443662431505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,1024,0.012821333275900947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,768,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,768,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,256,0.0032239999208185407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,512,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,512,0.011535111400816174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,256,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,256,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,128,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,256,0.010737777584128909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,2560,0.01789866718981001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,64,0.0028435554769304064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,128,0.006515555497672822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2560,32,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,128,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,65536,0.13562844859229192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,64,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,1024,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,32,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,768,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,65536,0.18378310733371309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,65536,0.07323200172848172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,16384,0.040928890307744346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2560,512,0.007126222054163615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,51200,0.1446977721320258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,51200,0.05882133377922905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,65536,0.21720533900790742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,12288,0.031762666172451444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,16384,0.02197955548763275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,51200,0.1703253322177463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,16384,0.04919377631611294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,10240,0.025096888343493145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,16384,0.0624835557407803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,12288,0.03804444604449802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,12288,0.01828977796766493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,8192,0.01947377787695991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,12288,0.049914668003718056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,128,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,10240,0.016749334004190233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,7168,0.018381334013409086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,10240,0.03289511137538486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,8192,0.013759111364682516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,10240,0.04321600000063578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,8192,0.027679999669392902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,6144,0.01601777805222405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,7168,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,8192,0.03658933440844218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,7168,0.025417778227064345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,5120,0.01352266636159685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,6144,0.011355555719799466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,51200,0.12864355246225992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,4096,0.011614222493436603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,6144,0.022516444325447083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,5120,0.010402667025725046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,6144,0.029896888468000624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,3584,0.010672888822025724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,5120,0.01907111042075687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,4096,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,5120,0.026828444666332666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,3072,0.009684444301658207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,4096,0.01606755620903439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,3584,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,4096,0.023600000474188063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,2560,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,3072,0.009016888837019602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,3584,0.0207004447778066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,3072,0.014035556051466199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,2048,0.0070222218831380205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,3072,0.019666666785875957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,2560,0.008432000047630733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,2560,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,1536,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,2560,0.017287999391555786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,2048,0.007958222594526079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,2048,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,1024,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,2048,0.015807999504937064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,1536,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,1536,0.009015999734401703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,768,0.004523555437723796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,1536,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,7168,0.03342577815055847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,1024,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,512,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,1024,0.012432888978057437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,768,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,768,0.011737777955002256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,512,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,512,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,256,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,3584,0.014781332678265043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,512,0.01106844428512785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,256,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,256,0.006302222195598815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,256,0.010312000082598792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,64,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,128,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,128,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,32,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,128,0.008347555167145198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,65536,0.1202053361468845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,64,0.00600000015563435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,32,0.005979555762476391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,1024,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,51200,0.08589777681562637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,65536,0.14363466368781194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,65536,0.07201066944334242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,2048,768,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,16384,0.03100799851947361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,65536,0.20483822292751738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,51200,0.05782755878236559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,51200,0.11334844430287679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,12288,0.022819555468029443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,16384,0.022279111875428095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,16384,0.038689778910742864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,51200,0.16313066747453478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,10240,0.019128888845443726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,16384,0.05940710835986667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,12288,0.01815022197034624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,12288,0.03054311209254795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,8192,0.015440000428093804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,12288,0.04741866721047295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,10240,0.015878222054905362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,7168,0.014118222726715935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,10240,0.026509332987997267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,10240,0.041032890478769936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,6144,0.012070222033394707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,7168,0.012072000238630505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,8192,0.03519999980926514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,7168,0.0206666671567493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,5120,0.010777778095669217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,7168,0.03189333279927572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,6144,0.011153777440388998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,6144,0.018244443668259513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,4096,0.009240888886981541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,6144,0.028655999236636694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,5120,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,5120,0.015795555379655626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,3584,0.008375110725561777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,5120,0.025592888394991558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,4096,0.009678222239017487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,2048,128,0.003142222141226133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,3072,0.007382222347789341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,4096,0.02223555578125848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,3584,0.012924444344308643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,3584,0.021116443806224402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,2560,0.006360000206364526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,3072,0.008810666700204214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,3072,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,8192,0.013406222065289816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,3072,0.018569777409235638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,8192,0.022619555393854778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,2560,0.00832622249921163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,2560,0.010230222509966956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,1536,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,2048,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,2048,0.009288888838556077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,2048,0.015474667151769003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,1024,0.003863999827040566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,1536,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,1536,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,1536,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,768,0.0034942221310403612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,1024,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,1024,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,1024,0.012447111308574677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,3584,0.009370666411187913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,512,0.0031715554909573663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,768,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,768,0.0120666664507654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,512,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,256,0.002611555573013094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,512,0.006071110980378256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,512,0.011086222198274402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,256,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,2560,0.016935111747847665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,256,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,128,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,256,0.010415111151006486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,64,0.002287111141615444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,128,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,32,0.0025048889219760895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,128,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,128,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,64,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,4096,0.013605333036846586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,32,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,51200,0.06519289149178399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1536,768,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,65536,0.13677510950300428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,65536,0.20232800642649332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,51200,0.058515555328793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1536,2048,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,16384,0.023960888385772705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,51200,0.10695822371376885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,12288,0.01959377858373854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,51200,0.1608817842271593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,16384,0.021465778350830078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,10240,0.016398222910033334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,16384,0.03769422239727444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,12288,0.01702044407526652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,16384,0.058543999989827476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,12288,0.029740442832310993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,8192,0.01401600076092614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,12288,0.04605422086185879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,10240,0.015490666031837463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,10240,0.025847999585999384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,7168,0.01276088919904497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,65536,0.08015200164582995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,10240,0.04030222362942166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,8192,0.021137777302000258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,8192,0.013063111239009433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,6144,0.01110666659143236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,8192,0.033992889854643084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,7168,0.012068444656001197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,5120,0.009766221874290043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,65536,0.0730817781554328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,7168,0.019315555691719055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,4096,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,7168,0.030262221892674763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,6144,0.017290666699409485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,6144,0.027169777287377253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,3584,0.0075084442893664045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,5120,0.015462223026487561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,4096,0.009381333159075843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,5120,0.02420355545149909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,3072,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,4096,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,4096,0.02146044373512268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,3584,0.012079111403889127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,2560,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,3584,0.01978577839003669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,3072,0.008647999829716152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,3072,0.017702221870422363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,2048,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,2560,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,2560,0.010248888697889116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,2560,0.016240888171725802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,1536,0.004595555365085602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,2048,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,2048,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,2048,0.015191111299726697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,6144,0.011055111057228513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,1024,0.0038693332009845306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,1536,0.0074346669846110875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,1536,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,768,0.0033679999825027254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,5120,0.010283555421564314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,1536,0.01352533366945055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,1024,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,1024,0.006917333437336816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,512,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,768,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,1024,0.0124195557501581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,256,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,768,0.006483555667930179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,3584,0.008981333010726506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,768,0.011722666521867117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,512,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,256,0.006108444597986009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,512,0.010729778144094678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,256,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,64,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,256,0.009713778065310584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,128,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,32,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,128,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,65536,0.059078223175472684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,64,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,32,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,65536,0.09682133462693955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,65536,0.07155644231372409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,51200,0.0518124434683058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,65536,0.20161244604322645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,51200,0.07656533188290067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,51200,0.05634755558437771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,512,0.00656444455186526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,1024,128,0.002238222294383579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,51200,0.15937956174214682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,1024,3072,0.011134222149848938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,12288,0.016484444340070088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,16384,0.02068444424205356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,16384,0.027817779117160376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,10240,0.015684444043371413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,16384,0.05803822146521675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,12288,0.016567111015319824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,128,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,12288,0.04553866717550489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,10240,0.015232889188660515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,10240,0.019343111250135634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,10240,0.03990133272276984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,8192,0.016188444362746347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,7168,0.011751111182901593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,8192,0.03382666574584113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,6144,0.010344888601038191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,7168,0.014937778313954672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,7168,0.011762667033407422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,7168,0.030180444320042927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,16384,0.01889600025282966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,5120,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,6144,0.011024889018800525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,6144,0.027181333965725366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,5120,0.012178666889667511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,5120,0.010011555420027839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,4096,0.007595555649863348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,5120,0.024152888192070857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,12288,0.022272000710169475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,3584,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,8192,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,4096,0.009373333719041612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,3072,0.00609422226746877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,4096,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,3584,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,4096,0.021027555068333943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,2560,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,3584,0.019655111763212416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,8192,0.012764444781674279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,3072,0.008644444247086843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,2048,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,3072,0.017941332525677152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,2560,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,2560,0.008627555436558193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,1536,0.00416711096962293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,2560,0.016504888733228047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,2048,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,2048,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,1024,0.00351200004418691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,2048,0.01520088811715444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,6144,0.013552000125249227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,1536,0.006909333169460297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,768,0.003167111012670729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,1536,0.013455111119482251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,1024,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,1024,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,512,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,1024,0.012112889024946423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,768,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,768,0.011736888852384357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,3584,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,512,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,512,0.01108888867828581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,3072,0.009310222334331935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,128,0.0021582223061058256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,256,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,256,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,256,0.010599111517270407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,64,0.002157333410448498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,128,0.0063075555695427795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,32,0.002163555473089218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,1536,0.007352888584136963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,64,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,65536,0.04501510990990532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,32,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,768,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,51200,0.03773422373665704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,65536,0.07111110952165392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,768,256,0.002471111093958219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,512,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,65536,0.20065422852834067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,16384,0.017840888765123155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,51200,0.05622844563590156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,51200,0.07479555739296807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,12288,0.01402311192618476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,51200,0.15913155343797472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,16384,0.027104000250498455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,16384,0.01994400057527754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,16384,0.057335111829969615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,768,128,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,12288,0.016414221790101793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,12288,0.02163822286658817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,128,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,8192,0.013108444710572561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,12288,0.0451164444287618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,10240,0.014934221903483072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,10240,0.018872888551818002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,7168,0.011367110742463006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,10240,0.03924977779388428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,8192,0.012481778032249875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,8192,0.016308445069524977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,6144,0.010371555884679159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,65536,0.09469244215223525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,7168,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,7168,0.014709333578745524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,5120,0.008901333643330468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,7168,0.029335111379623413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,6144,0.010760889285140567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,6144,0.013443555268976422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,4096,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,6144,0.026553778184784785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,5120,0.010044444766309526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,5120,0.02365600069363912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,4096,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,3584,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,10240,0.012619555824332766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,4096,0.01036800030204985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,3072,0.006224888894293044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,4096,0.020868443780475195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,3584,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,3584,0.010266666611035665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,2560,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,3072,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,3072,0.00903200027015474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,3072,0.017567111386193168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,2048,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,8192,0.032608889871173434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,2560,0.008644444247086843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,1536,0.004223111189073986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,2560,0.016176000237464905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,2048,0.007678222325113084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,2048,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,1024,0.003499555504984326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,2048,0.014687110980351767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,1536,0.0073520003093613525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,1536,0.006818666640255187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,5120,0.011925333076053195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,768,0.0031973332580592898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,1536,0.013094222380055321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,1024,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,512,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,1024,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,768,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,768,0.00572622236278322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,1024,0.011756444142924415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,256,0.002536000063021978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,768,0.011051555474599203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,512,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,128,0.002191999927163124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,512,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,3584,0.019267555740144517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,256,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,256,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,64,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,512,32,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,256,0.009649777577983009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,128,0.005975111077229182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,65536,0.030244443151685927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,2560,0.008042666647169325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,128,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,64,0.004961777892377642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,51200,0.026403556267420452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,32,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,65536,0.06990755266613431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,65536,0.09327644771999782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,16384,0.012181333369678922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,51200,0.05401866634686788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,51200,0.07340444458855523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,65536,0.20065777831607393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,12288,0.01037333326207267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,16384,0.019632889164818656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,51200,0.1589911116494073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,16384,0.02632355524433984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,10240,0.009453333086437648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,16384,0.056102222866482206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,12288,0.016184000505341422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,12288,0.02094399929046631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,8192,0.00831822223133511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,10240,0.014827556080288358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,12288,0.04384533233112759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,10240,0.01830666595035129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,7168,0.007763555480374231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,512,0.010390222072601318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,10240,0.03813866774241129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,8192,0.012423999607563019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,6144,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,8192,0.015843555331230164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,8192,0.03220266766018338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,7168,0.011728888584507836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,512,128,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,5120,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,7168,0.014506666196717156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,6144,0.010686222049925061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,4096,0.006912888752089606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,7168,0.029161777761247423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,6144,0.013106666505336761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,5120,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,6144,0.02623288830121358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,3584,0.006423111177153057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,5120,0.011756444142924415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,4096,0.009303111169073317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,3072,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,3584,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,4096,0.020471110939979553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,3584,0.009846222069528367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,2560,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,3584,0.019239111079110038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,3072,0.008572444319725037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,3072,0.008972444468074376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,2048,0.004500444564554426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,3072,0.01751377847459581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,2560,0.008053333395057255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,2560,0.008461332983440822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,1536,0.0038853333228164246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,2560,0.015842666228612263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,2048,0.0076657773719893555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,2048,0.007608888877762689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,1024,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,2048,0.01479200025399526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,1536,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,1536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,768,0.0028728888266616394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,1536,0.01310222182008955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,1024,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,1024,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,512,0.0025191110455327565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,1024,0.011779555843936073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,5120,0.0236799998415841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,768,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,4096,0.01032266683048672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,256,0.002500444443689452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,768,0.010768888725174798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,512,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,512,0.006411555740568373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,128,0.002102222293615341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,256,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,512,0.010679999987284342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,256,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,64,0.002069333361254798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,256,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,256,32,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,128,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,65536,0.024912888805071514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,128,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,64,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,51200,0.018358222312397428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,65536,0.0695813364452786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,65536,0.09277777539359199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,16384,0.011459555890825061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,65536,0.20146577888064912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,51200,0.05306666758325365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,51200,0.07300444444020589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,12288,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,768,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,16384,0.02586311101913452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,51200,0.15857156117757162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,10240,0.008611555728647444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,16384,0.05592711104287041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,12288,0.016368889146380954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,12288,0.020695999264717102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,8192,0.007671111159854465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,12288,0.043531556924184166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,10240,0.018165333403481376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,128,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,10240,0.037998222642474704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,7168,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,8192,0.015576000014940897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,256,32,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,8192,0.031874666611353554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,7168,0.01147822207874722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,6144,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,7168,0.014430221584108142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,7168,0.028972443607118394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,6144,0.010741333166758219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,6144,0.013125333521101209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,6144,0.02608977754910787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,4096,0.006319111006127463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,16384,0.01955200069480472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,5120,0.01163111130396525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,5120,0.009848888549539778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,5120,0.023344000180562336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,4096,0.009295110901196798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,4096,0.010103111465771994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,4096,0.02018933329317305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,10240,0.014885332849290637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,3072,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,3584,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,3584,0.008960000342792934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,2560,0.004860444615284602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,8192,0.01239733315176434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,3584,0.018927110566033255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,3072,0.008448888858159384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,3072,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,2048,0.004515555583768421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,2560,0.008043555749787224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,3072,0.016903110676341586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,2560,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,1536,0.0038515557017591265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,2560,0.015811555915408664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,5120,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,2048,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,1024,0.003169777699642711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,2048,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,1536,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,2048,0.014482667048772177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,1536,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,1536,0.012764444781674279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,1024,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,1024,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,3584,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,512,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,1024,0.012030222349696688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,768,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,768,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,256,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,512,0.0063528890411059065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,768,0.01073688848151101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,512,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,512,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,256,0.006215999937719769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,256,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,64,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,256,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,32,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,128,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,65536,0.023581332630581323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,128,0.007333333293596904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,64,0.004945777770545748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,128,32,0.005052444421582752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,16384,0.008640888664457533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,65536,0.0924346645673116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,768,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,12288,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,10240,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,51200,0.0725733306672838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,8192,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,16384,0.025479111406538222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,7168,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,12288,0.020658666888872784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,10240,0.017990221579869587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,8192,0.015455999308162265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,5120,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,128,128,0.0021280000607172647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,7168,0.014334221680959066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,4096,0.005993777679072485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,3584,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,6144,0.012838222086429596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,128,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,3072,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,2560,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,5120,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,2048,0.0042151109211974675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,4096,0.010399999717871347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,3584,0.009387555221716562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,51200,0.018199111024538677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,1024,0.0031644445326593188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,3072,0.009017777939637503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,768,0.002797333316670524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,512,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,2560,0.00833777752187517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,2048,0.007322666545708974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,256,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,128,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,1536,0.00665866666369968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,64,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,1024,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,768,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,32,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,512,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,256,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,128,0.004951111144489712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,51200,0.03126577867401971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,64,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,16384,0.013083555632167392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,64,32,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,6144,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,10240,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,65536,0.09200177590052287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,8192,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,51200,0.07225422064463298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,16384,0.025594666600227356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,7168,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,12288,0.020672889219390023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,6144,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,5120,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,10240,0.0181697772608863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,4096,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,8192,0.015488000379668342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,64,1536,0.0036346668170558084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,3584,0.005873777800136142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,7168,0.014149333039919535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,3072,0.005532444351249271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,6144,0.013109332985348172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,5120,0.011418666276666852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,4096,0.010108444425794814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,2048,0.0042453333735466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,3584,0.009654222263230218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,1536,0.003916444463862313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,3072,0.00869599978129069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,1024,0.0031537777847713898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,2560,0.008226667013433244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,65536,0.018375999397701688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,768,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,512,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,2048,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,256,0.0021608887861172357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,1536,0.006672888994216919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,12288,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,768,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,64,0.001810666587617662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,512,0.0052844443255000645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,32,0.001964444501532449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,256,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,128,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,64,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,32,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,2560,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,768,32,128,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,16384,0.7529484430948893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,16384,0.7402293417188855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,16384,0.2994355625576443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,12288,0.6130293210347494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,12288,0.543184863196479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,51200,2.3933768802218967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,768,32,1024,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,12288,0.22787290149264863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,10240,0.44300444920857746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,10240,0.46315823660956484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,8192,0.36580801010131836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,51200,2.522704866197374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,10240,0.19848977194892037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,8192,0.16130577193366155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,7168,0.36531999376085067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,8192,0.3656293286217584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,6144,0.2998853259616428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,7168,0.3200799889034695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,6144,0.1205022202597724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,6144,0.27821866671244305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,4096,0.17635644806755912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,5120,0.10623288816875881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,7168,0.14037511083814833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,5120,0.23271200391981336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,3584,0.15425688690609402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,4096,0.08857155508465236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,4096,0.1876222160127428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,3072,0.1314728922314114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,2560,0.13835733466678196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,3584,0.07589511076609294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,3072,0.06792266501320733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,3584,0.16389689180586073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,2048,0.10075555245081584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,3072,0.14545688364240858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,2560,0.05792533026801216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,1536,0.07064355744255914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,2560,0.12156355381011963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,2048,0.051185776789983116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,1024,0.0502106679810418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,2048,0.1000515553686354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,1536,0.0401777790652381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,768,0.03999200132158067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,1536,0.07979822158813477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,1024,0.030997332599427965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,5120,0.30969511138068306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,256,0.02242044442229801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,768,0.028299556838141546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,1024,0.06037777662277222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,512,0.025968889395395916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,128,0.01864177816443973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,768,0.04871911141607496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,512,0.04236355423927307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,64,0.015783111254374187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,51200,0.9924159579806857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,32,0.016515556308958266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,128,0.023363555471102398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,256,0.03738133443726434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,128,0.03494488861825731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,64,0.0348417792055342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,65536,32,0.034712887472576566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,65536,512,0.029683556821611192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,65536,256,0.023736889163653057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,16384,0.5323484208848741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,16384,0.5803910891215006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,51200,1.9242515563964844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,16384,0.2430222299363878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,51200,0.7956319914923774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,65536,1.0302453570895724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,65536,2.4920096927218967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,12288,0.39649510383605957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,51200,1.9837040371365016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,65536,2.5100534227159286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,12288,0.18351822429233125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,10240,0.4454524252149794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,8192,0.27737779087490505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,12288,0.42930489116244847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,7168,0.2764497862921821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,10240,0.5113813082377116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,8192,0.12639466921488443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,8192,0.2848728762732612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,10240,0.15660266081492105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,5120,0.1727235582139757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,6144,0.2600088914235433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,6144,0.09760622183481853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,7168,0.25800977812872994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,7168,0.11549777454800075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,6144,0.22202577855851915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,4096,0.17794133557213676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,5120,0.1819902261098226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,5120,0.08397422234217326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,3584,0.15659199820624456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,4096,0.06752088997099134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,3072,0.11451110574934219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,4096,0.14944888485802546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,2560,0.09869688749313354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,3584,0.061647110515170626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,2048,0.07843022214041816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,2560,0.04714577727847629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,1536,0.06307111183802287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,2560,0.09613333145777385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,2048,0.04087466663784451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,2048,0.08522844314575195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,1024,0.04973333411746555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,1536,0.0325546662012736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,768,0.036836445331573486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,1536,0.06343199809392293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,1024,0.025690666503376428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,512,0.02679466704527537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,1024,0.048884444766574435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,768,0.023080888721677992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,256,0.019357333580652874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,768,0.03968355390760634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,512,0.02124533388349745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,128,0.01484444406297472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,512,0.0332275562816196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,256,0.019347555107540555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,64,0.012921777864297232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,256,0.029478222131729126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,128,0.01921244462331136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,51200,32,0.013073777986897362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,128,0.028132445282406274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,3584,0.13238488303290472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,64,0.028000000450346205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,3072,0.11254933145311143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,51200,32,0.027600000301996868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,16384,0.1777120033899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,65536,0.7552017635769315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,65536,0.339430226220025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,51200,0.2711057662963867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,65536,0.788082652621799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,16384,0.18915111488766143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,51200,0.6319333182440864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,16384,0.08819378084606594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,10240,0.11790222591824001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,12288,0.14350667264726427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,12288,0.065702219804128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,51200,3072,0.05535555548138089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,8192,0.09682044718000625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,10240,0.0594995551639133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,10240,0.12056088447570801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,7168,0.09437777598698933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,6144,0.07427199681599934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,7168,0.04179644584655762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,7168,0.08592089017232259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,8192,0.09893955787022908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,6144,0.07546666595670912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,4096,0.055733336342705615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,51200,0.6093120045132107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,5120,0.03162666824128892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,5120,0.0637342201338874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,3584,0.04892888996336195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,4096,0.027493332823117573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,3072,0.043466667334238686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,12288,0.16599555810292563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,3584,0.02459999918937683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,2560,0.03607733382119073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,3584,0.046542220645480685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,2560,0.019644444187482197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,3072,0.041264888313081526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,2560,0.03571288784344991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,2048,0.028004444307751123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,8192,0.04632000128428141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,1536,0.022083555658658344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,2048,0.017714665995703805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,1024,0.0158915552828047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,5120,0.06806488831837972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,2048,0.02983022067281935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,1536,0.015085332923465304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,6144,0.03677333394686381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,768,0.013152889079517789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,512,0.010293333066834344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,1536,0.024636444118287828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,1024,0.012440889245933957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,768,0.011507555842399597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,256,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,1024,0.01958933307064904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,512,0.010760000182522668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,768,0.016548444827397663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,128,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,512,0.013771555489963956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,256,0.010053333308961656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,64,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,4096,0.05276799864239163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,16384,32,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,128,0.010009778042634329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,128,0.012335999972290464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,64,0.01240533341964086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,32,0.012265777422322167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,16384,3072,0.02239466706911723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,65536,0.45752620697021484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,51200,0.35730399025811094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,65536,0.24683022499084473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,65536,0.5953066613939073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,51200,0.1939368910259671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,12288,0.09352888663609822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,51200,0.4668897522820367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,16384,0.06434400214089288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,16384,0.1568897830115424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,12288,0.0488737788465288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,16384,256,0.012620444099108378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,12288,0.11104889048470391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,10240,0.04279377725389269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,8192,0.0788737800386217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,10240,0.09290933609008789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,7168,0.07097422414355807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,8192,0.03470755616823832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,6144,0.05948978000217014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,7168,0.03057510985268487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,8192,0.07601688967810737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,5120,0.052928000688552856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,7168,0.06786488824420504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,6144,0.027595554788907368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,4096,0.03941599859131707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,6144,0.05979910824033949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,5120,0.02402666707833608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,3584,0.03122399912940131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,5120,0.051142222351498075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,4096,0.020979555116759405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,3072,0.02966844373279148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,4096,0.041787554820378624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,3584,0.019447111421161227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,3584,0.036730666955312095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,3072,0.018735999862353008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,2048,0.01994400057527754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,3072,0.03282311227586534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,10240,0.0867040024863349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,1536,0.016735111673672993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,2560,0.028696888022952612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,2048,0.014538667268223233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,1024,0.01181688904762268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,1536,0.012662221988042196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,2048,0.02421333392461141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,768,0.009649777577983009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,1536,0.020224889119466145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,1024,0.010950222611427307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,512,0.007627555893527136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,768,0.009724444813198512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,256,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,512,0.009378666679064432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,768,0.013748444616794586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,128,0.004864888886610667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,512,0.011689778003427716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,256,0.008638222184446123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,64,0.004280000097221798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,128,0.008952889177534316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,16384,0.13226844204796687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,32,0.004534222185611725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,128,0.01033866653839747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,12288,2560,0.026494221554862127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,64,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,32,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,12288,2560,0.01625866691271464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,65536,0.5091386901007758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,65536,0.2285217709011502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,51200,0.40017422040303546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,16384,0.13184266620212132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,65536,0.5029786427815756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,1024,0.016191111670600045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,51200,0.17842933866712782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,51200,0.4049475457933214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,10240,0.09704266654120551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,12288,0.10313244660695393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,16384,0.05985955397288004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,12288,0.04693866769472758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,16384,0.12503288851843938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,8192,0.07473866807089911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,12288,0.09355822536680435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,7168,0.06926844517389934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,12288,256,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,10240,0.040236443281173706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,8192,0.03293155630429586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,10240,0.0793822209040324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,7168,0.02974489000108507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,5120,0.048751112487581044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,8192,0.06528444422615899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,6144,0.026170666019121807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,7168,0.057982219590081104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,4096,0.03799644443723891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,6144,0.05109599894947476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,3584,0.03379911184310913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,5120,0.023253333237436082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,5120,0.043599112166298755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,4096,0.02016799979739719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,4096,0.03568089008331299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,3584,0.018569777409235638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,2560,0.025447110335032146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,3072,0.017343999611006845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,3584,0.032010667853885226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,2048,0.02104088829623328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,3072,0.02839822239345974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,2560,0.015437333120240105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,2048,0.013963555296262106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,2560,0.025057777762413025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,1024,0.012355555262830524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,2048,0.021359110871950786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,1536,0.012170666621790992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,1536,0.017880888448821176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,768,0.010384000009960597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,1024,0.01033333357837465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,512,0.007879999776681265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,1024,0.01422044469250573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,768,0.00965244405799442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,6144,0.05617955658170912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,256,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,768,0.012079999678664737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,512,0.008987555901209513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,128,0.004859555512666702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,512,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,256,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,64,0.0041786668201287585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,10240,128,0.008660444782839881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,32,0.004521777646409141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,3072,0.030071109533309937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,128,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,64,0.009441778063774109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,32,0.009371555513805812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,10240,1536,0.016199999385409884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,65536,0.3639404508802626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,51200,0.2641937732696533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,16384,0.110390219423506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,65536,0.41066400210062665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,12288,0.07758044534259372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,51200,0.319942209455702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,16384,0.10261688629786174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,10240,0.06509866979387072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,12288,0.03752888904677497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,12288,0.07808888620800443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,8192,0.050716443194283374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,10240,0.03276177909639146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,10240,256,0.00977244476477305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,10240,0.06625155607859294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,7168,0.04712888929578993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,6144,0.0406257775094774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,8192,0.026590221458011206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,8192,0.05453866720199585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,7168,0.024617777930365667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,5120,0.03367644548416138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,7168,0.04799377918243408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,6144,0.021741333934995864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,4096,0.02737688852681054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,6144,0.04284177886115181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,5120,0.01959022217326694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,3584,0.02446577780776554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,65536,0.17968800332811144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,5120,0.03665688965055678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,4096,0.0174346665541331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,3072,0.021337777376174927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,4096,0.029835555288526747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,2560,0.018106665876176622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,3584,0.016710221767425537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,3072,0.014884443746672736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,51200,0.14243911372290716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,3584,0.02748799986309475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,16384,0.05060355530844795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,2560,0.01349066694577535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,2048,0.012135110795497894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,2560,0.02127911150455475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,1024,0.009882666998439366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,1536,0.010416000253624385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,2048,0.018329777651362948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,768,0.008278222547637092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,1536,0.015504888362354703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,1024,0.009087111386987898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,1024,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,768,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,256,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,768,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,512,0.008210666477680206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,512,0.009391110804345872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,256,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,256,0.009024000002278222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,64,0.0038977778620190094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,8192,128,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,32,0.00388355553150177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,128,0.00869066682126787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,2048,0.01570133368174235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,64,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,1536,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,32,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,65536,0.3408942222595215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,512,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,51200,0.2984844313727485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,65536,0.16936355166965059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,65536,0.37259822421603733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,16384,0.08684621916876899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,12288,0.0726062191857232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,8192,128,0.004176889028814104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,51200,0.13517955938975015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,16384,0.04708177844683329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,51200,0.3110568788316515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,16384,0.09094399876064724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,12288,0.03640266590648227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,8192,0.04473333226309883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,12288,0.07468089130189684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,10240,0.032524443335003324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,7168,0.04284888837072584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,10240,0.05950755543178982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,8192,0.027116444375779893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,6144,0.03247822324434916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,8192,0.048376000589794584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,8192,3072,0.024455999334653217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,7168,0.02383199996418423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,5120,0.028720888826582167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,7168,0.043975111511018544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,4096,0.022673777408070032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,6144,0.022032888399230108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,3584,0.020234665936893888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,6144,0.03815111186769273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,5120,0.019346666004922655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,3072,0.018997333116001554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,4096,0.017288888494173687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,5120,0.03307111064592997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,3584,0.01591377788119846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,2560,0.015404444601800708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,4096,0.027592000034120347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,2048,0.014257777068350049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,3584,0.02520177761713664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,2560,0.012967999610635968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,3072,0.02223199937078688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,1536,0.011390222443474663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,2048,0.011749332977665795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,1024,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,2048,0.01682400041156345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,1536,0.010335110955768162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,768,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,1536,0.014333332578341166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,1024,0.00904088881280687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,512,0.005842666659090254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,1024,0.011344888971911537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,768,0.008377778033415476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,256,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,768,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,512,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,128,0.003842666745185852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,512,0.008784888519181145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,64,0.0035217776894569397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,128,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,32,0.003552000141806073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,128,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,64,0.00831288927131229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,3072,0.014357333381970724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,32,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,65536,0.23610133594936797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,2560,0.01957066688272688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,51200,0.1818693346447415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,7168,10240,0.06048533651563856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,16384,0.06160710917578804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,65536,0.13145421610938177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,65536,0.31732267803615993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,12288,0.05610133541954888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,51200,0.10395822260114883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,51200,0.2530826727549235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,10240,0.04618311259481642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,16384,0.07983644141091241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,16384,0.04010933306482103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,12288,0.028064890040291682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,8192,0.03827288746833801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,7168,0.034317334493001304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,12288,0.061407109101613365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,10240,0.02551555633544922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,6144,0.028276443481445312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,8192,0.02068711154990726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,7168,256,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,10240,0.052368889252344765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,8192,0.043213334348466664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,7168,0.019123555885420907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,4096,0.019360888335439894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,6144,0.017486222916179232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,7168,0.0381022228135003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,5120,0.016192000773217943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,6144,0.03394488824738397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,5120,0.02921600143114726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,4096,0.014774221513006421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,3072,0.014195554786258273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,4096,0.024803555674023096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,3584,0.022613333331214056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,2560,0.012874667015340594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,3072,0.012585777375433179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,3072,0.020072888996866014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,2048,0.010947555303573608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,2560,0.010716444088353051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,2560,0.017838221457269456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,1536,0.009286222358544668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,2048,0.015432000160217285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,7168,256,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,1024,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,1536,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,1536,0.012513777448071374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,5120,0.02520711057715946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,768,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,1024,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,512,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,3584,0.017217778497272067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,768,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,256,0.0038604442444112566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,768,0.009334222310119206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,512,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,3584,0.014441778262456259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,512,0.008284444610277811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,256,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,64,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,256,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,128,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,32,0.0031635556370019913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,128,0.007672000262472365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,64,0.007663110891977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,32,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,65536,0.25807200537787545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,6144,2048,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,16384,0.06905066967010498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,51200,0.22789245181613496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,65536,0.1233324474758572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,65536,0.2698853280809191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,51200,0.09767555528216892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,12288,0.05557511250178019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,16384,0.03497155507405599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,10240,0.044510222143597074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,16384,0.06878933641645643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,12288,0.027768888407283362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,8192,0.037187556425730385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,12288,0.05334844523006015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,10240,0.024442666106753882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,6144,128,0.0032595555401510666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,7168,0.03318488929006789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,10240,0.05298577745755514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,8192,0.02049333353837331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,8192,0.037476443582110934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,7168,0.01886666648917728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,5120,0.023876445160971746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,6144,0.017312000195185345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,7168,0.03347644540998671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,4096,0.0196426659822464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,6144,0.030244443151685927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,5120,0.015612444943851896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,3584,0.01769244505299462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,5120,0.02621777852376302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,3072,0.015626667274369132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,4096,0.021872889664438035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,3584,0.013456888496875763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,2560,0.013468444347381592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,3584,0.020046222541067336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,3072,0.01203733351495531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,2048,0.011349333657158745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,3072,0.018145778112941317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,2560,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,51200,0.21161954932742646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,1536,0.00944977750380834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,2560,0.015990222493807476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,1024,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,2048,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,1536,0.008972444468074376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,2048,0.01352266636159685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,768,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,1024,0.008114666574531132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,1536,0.011174221833546957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,6144,0.028939555088678997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,512,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,1024,0.009411555197503831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,768,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,256,0.0042035554846127825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,128,0.003519111209445529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,768,0.008456000023418004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,512,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,256,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,64,0.003152888889114062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,512,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,5120,32,0.0032302221904198327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,256,0.007350222104125553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,128,0.006994666324721442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,128,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,5120,4096,0.013727111121018728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,64,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,65536,0.21204977565341523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,51200,0.19007021851009795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,16384,0.05676177475187513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,65536,0.1166871123843723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,65536,0.22732355859544542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,51200,0.17629067103068033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,51200,0.09370133611891006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,12288,0.04682755470275879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,16384,0.03388088941574097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,6144,1024,0.010431999961535135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,10240,0.038542220989863075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,16384,0.05804889069663154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,8192,0.03204444381925795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,12288,0.027064000566800434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,7168,0.027457776996824477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,10240,0.024853333830833435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,12288,0.044601778189341225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,8192,0.020765332712067496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,10240,0.03837777839766608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,6144,0.02292266653643714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,5120,0.019521777828534443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,8192,0.03172355559137132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,6144,0.016732444365819294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,7168,0.028760890165964764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,4096,0.016520889268981088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,5120,0.014834667245546976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,6144,0.02556444373395708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,3584,0.014813333749771118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,5120,0.022630222969584998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,3072,0.012921777864297232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,4096,0.013056000073750814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,3584,0.011778666741318174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,4096,0.01904355486234029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,5120,32,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,2560,0.011505777637163797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,2048,0.009694221946928237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,3072,0.011057777537239922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,2560,0.010060444474220276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,2048,0.009268444445398118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,1536,0.007868444754017724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,2048,0.011702222128709158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,1536,0.008617777791288164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,1024,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,1536,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,768,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,1024,0.007706666986147563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,1024,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,7168,0.019003555178642273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,512,0.004504000147183736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,768,0.007334222396214803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,256,0.0035537779331207275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,768,0.00794488853878445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,512,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,128,0.0035582222044467926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,512,0.007073777417341868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,256,0.006918222126033571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,64,0.0028444443725877335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,4096,128,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,3584,0.01751733322938283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,4096,32,0.0031226666437255014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,128,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,3072,0.015821332732836407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,64,0.006693333387374878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,2560,0.013471999930010902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,65536,0.18140267001258004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,32,0.006672888994216919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,51200,0.14394577344258627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,16384,0.05479200018776787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,65536,0.20997956064012316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,51200,0.09267822239134048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,12288,0.036023110151290894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,16384,0.032836443848080106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,10240,0.03030666708946228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,16384,0.05624177720811632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,8192,0.0255031105544832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,12288,0.04284888837072584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,10240,0.023355555203225877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,7168,0.022171555293930903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,8192,0.019711110326978896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,4096,256,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,8192,0.030805332793129817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,6144,0.020114666885799833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,7168,0.01807822287082672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,7168,0.028103109863069322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,5120,0.01645955608950721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,6144,0.016304888659053378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,4096,0.014236445228258768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,6144,0.024774221910370722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,5120,0.014748444159825643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,3584,0.01272266689274046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,5120,0.021891555852360193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,4096,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,65536,0.11553777588738336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,4096,0.018746667438083224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,3584,0.011834666960769229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,3584,0.016906667086813185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,51200,0.16774133841196695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,3072,0.015187554889255099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,12288,0.026465778549512226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,2048,0.00865955568022198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,2560,0.00999466660949919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,1536,0.007244444555706448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,2560,0.013434666726324292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,10240,0.03733333283000522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,2048,0.00924711094962226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,1024,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,2048,0.011448000040319232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,1536,0.008366222182909647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,768,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,1536,0.010008888940016428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,1024,0.007672000262472365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,512,0.003882666842805015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,768,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,768,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,256,0.003245333416594399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,512,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,512,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,128,0.0028479999552170434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,3072,0.01111644423670239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,256,0.006680000159475539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,64,0.0025635556214385563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,256,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,3072,0.01094133324093289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3584,128,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,32,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,128,0.0063457778758472866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,65536,0.16327910953097874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,64,0.006607999818192587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,32,0.006414222220579783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,51200,0.1280871099895901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,65536,0.07605955335828993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,65536,0.18153511153327095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,16384,0.04336533281538221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,51200,0.06004977888531155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,12288,0.03376266691419814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,51200,0.13798844814300537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,16384,0.023019555542204116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3584,1024,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,10240,0.02847555610868666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,8192,0.023346667488416035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,10240,0.016857778032620747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,10240,0.03135822216669718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,7168,0.021957332889238994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,8192,0.014459555347760519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,8192,0.025918222135967676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,6144,0.01829333272245195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3584,2560,0.009903111391597325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,7168,0.013430222041077085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,5120,0.015801777442296345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,7168,0.023784889115227595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,6144,0.0119991112086508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,6144,0.021184888150956895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,4096,0.013059555656380124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,5120,0.011751111182901593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,3584,0.012004444168673621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,5120,0.018981332580248516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,4096,0.009755555954244401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,3072,0.010671111444632212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,4096,0.016027554869651794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,3584,0.00996088898844189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,2560,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,3584,0.014532445205582513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,16384,0.046889778640535146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,3072,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,2048,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,12288,0.019253333409627277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,1536,0.006787555499209299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,2560,0.00850399997499254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,3072,0.0129457778400845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,1024,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,12288,0.03622577918900384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,2048,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,2560,0.011434666812419891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,768,0.004599110947714912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,1536,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,512,0.0038737778862317405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,1024,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,2048,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,256,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,1536,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,1024,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,128,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,512,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,64,0.0024906665914588505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,256,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,512,0.006565333240562015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,3072,32,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,256,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,128,0.006052444378534953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,64,0.006084444622198741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,65536,0.16615733835432264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,32,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,51200,0.13661955462561712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,65536,0.07188444667392306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,65536,0.16928177409701875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,16384,0.04535377687878079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,12288,0.03489511211713155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,51200,0.13168711132473415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,16384,0.021903999977641638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,10240,0.028963555892308552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,12288,0.018017777138286166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,12288,0.03526488939921061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,10240,0.016064888901180692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,768,0.006671999891599019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,10240,0.030679110023710463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,8192,0.02344622214635213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,7168,0.020735111501481798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,3072,768,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,3072,128,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,8192,0.013941332697868347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,6144,0.018631999691327412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,7168,0.012726222475369772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,7168,0.02306933369901445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,6144,0.01146933353609509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,6144,0.0209840006298489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,4096,0.013526221944226159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,5120,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,5120,0.018181333939234417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,51200,0.057617776923709445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,4096,0.009704888694816166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,4096,0.015189333094490899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,16384,0.04516977734035916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,3584,0.009185777770148383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,3584,0.013911111487282647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,2560,0.009687999884287516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,3072,0.00867377801073922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,3072,0.01274577776590983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,2048,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,2560,0.008351999852392409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,2560,0.011232000258233813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,1536,0.006787555499209299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,8192,0.025642666551801894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,2048,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,5120,0.015712888704405892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,1024,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,1536,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,1536,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,768,0.004605333424276776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,1024,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,512,0.003536000019974179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,1024,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,3584,0.012143111063374413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,768,0.006663110935025745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,256,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,768,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,512,0.006310222049554189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,3072,0.010971555279360877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,256,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,64,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,256,0.005988444305128521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2560,128,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,128,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,65536,0.11605600516001384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,64,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,32,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,51200,0.09239733219146729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,65536,0.06810310814115736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,2048,0.009896000226338705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,16384,0.0378115541405148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,51200,0.05462933248943753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,51200,0.10293333397971259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,16384,0.020956445071432326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,12288,0.030737777551015217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,10240,0.026194666822751362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,12288,0.017647999856207106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,12288,0.028023110495673284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,128,0.0024968888610601425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2560,512,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,8192,0.021624000536070928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,10240,0.01595288846227858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,10240,0.02427111069361369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2560,32,0.002529777793420686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,8192,0.013388444152143268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,6144,0.017816889617178176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,7168,0.012096000214417776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,8192,0.02077155477470822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,5120,0.01605955594115787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,7168,0.01900888813866509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,4096,0.010975110861990186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,6144,0.011397333608733283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,6144,0.017142222987280954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,3584,0.010010666317409938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,5120,0.014803555276658801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,4096,0.009388444324334463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,65536,0.12879021962483725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,3072,0.008939555949634975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,4096,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,3584,0.008952000074916417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,2560,0.008224888808197444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,3072,0.008658666577604082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,2048,0.006744888921578725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,2560,0.008023999631404877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,16384,0.03559466534190708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,1536,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,2560,0.009546666509575313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,2048,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,1024,0.004549333204825719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,1536,0.007245333658324347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,7168,0.0207324441936281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,1536,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,1024,0.006859555426571105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,1024,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,768,0.006541333264774746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,512,0.0034933334423436057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,5120,0.01033866653839747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,512,0.006223111102978389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,512,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,256,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,3584,0.011796444654464722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,256,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,128,0.002493333278430833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,3072,0.010750222537252637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,256,0.0058711109062035876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,64,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,2048,128,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,128,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,2048,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,64,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,768,0.0038684445122877755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,65536,0.09639022085401748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,32,0.005640000104904175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,51200,0.08316000302632649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,16384,0.030642665094799463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,65536,0.06655555301242404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,65536,0.1276364458931817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,51200,0.0536702209048801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,2048,768,0.006436444405052397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,12288,0.02557955516709222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,51200,0.09932355748282538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,16384,0.020658666888872784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,12288,0.016901334126790363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,16384,0.034648001194000244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,8192,0.017934223016103108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,12288,0.027801778581407335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,7168,0.015903110305468243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,10240,0.023984000086784363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,8192,0.012785777449607849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,6144,0.014124444789356656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,8192,0.02013511127895779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,7168,0.011883555187119378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,5120,0.012088889049159156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,2048,32,0.00249422217408816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,7168,0.018393778138690524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,6144,0.010816888676749336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,4096,0.011056889262464313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,6144,0.016383111476898193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,3584,0.009948444035318162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,4096,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,5120,0.014549333188268872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,3072,0.007699555820888943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,4096,0.01238844460911221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,3584,0.008922666311264038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,2560,0.006650666809744305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,3584,0.01167555567291048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,2048,0.0058373332851462895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,3072,0.010741333166758219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,10240,0.021440888444582622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,1536,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,2560,0.00960888879166709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,2048,0.007595555649863348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,2048,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,10240,0.015241778559154935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,1024,0.0038488888078265717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,768,0.0035137778355015647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,1536,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,1536,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,512,0.0028773333049482773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,768,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,768,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,256,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,512,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,512,0.005965333431959152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,5120,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,128,0.002542222125662698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,256,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,64,0.0024622221373849446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,256,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1536,32,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,128,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,128,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,65536,0.06455466482374403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,3072,0.008282666405042013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,64,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,51200,0.05208177698983086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,32,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,16384,0.01958311100800832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,51200,0.05382400088840061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,65536,0.08126311169730292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,12288,0.016419554750124615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,1024,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,16384,0.024078221784697637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,16384,0.020054222808943856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,10240,0.01589777734544542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1536,1024,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,12288,0.016303999556435477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,12288,0.01942133406798045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,10240,0.014807111687130399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,7168,0.01202133297920227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,10240,0.017172444197866652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,8192,0.012447111308574677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,8192,0.014484445254007975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,6144,0.01094044413831499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,7168,0.011521778172916837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,7168,0.013463999662134381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,5120,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,6144,0.010727110836240979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,6144,0.012281777958075205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,4096,0.00868711123863856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1536,2560,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,5120,0.009751111268997192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,65536,0.06751200225618151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,3584,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,4096,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,3072,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,4096,0.009764444496896531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,51200,0.065065774652693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,2560,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,3584,0.009332444104883406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,3072,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,2048,0.005019555489222209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,3072,0.008625778059164682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,8192,0.013468444347381592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,2560,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,2560,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,2048,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,1024,0.0036159998012913596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,2048,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,1536,0.006979555719428592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,1536,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,768,0.003208000005947219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,1024,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,512,0.0028044444819291434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,768,0.006405333264006509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,768,0.0058115555180443665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,512,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,5120,0.011130666567219628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,512,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,128,0.0021768889079491296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,256,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,256,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,64,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,128,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,128,0.005092444519201915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,32,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,1536,0.004352000024583605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,65536,0.04730133215586344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,64,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,32,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,65536,0.06552800205018786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,65536,0.07954221963882446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,16384,0.016187555260128446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,51200,0.05173155665397644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,1024,1024,0.006137777947717243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,51200,0.06282311015658908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,12288,0.013521778086821238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,10240,0.012306667036480375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,16384,0.01905333333545261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,1024,256,0.0025235555238193935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,12288,0.01588800052801768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,8192,0.01389422184891171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,16384,0.023609777291615803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,12288,0.01891288823551602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,10240,0.01445777714252472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,1024,3584,0.008639111287064022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,8192,0.01236088905069563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,10240,0.016519111063745286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,6144,0.010461333725187512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,5120,0.00886311133702596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,8192,0.014122666584120857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,6144,0.01074488874938753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,4096,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,5120,0.009956444303194681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,6144,0.012016889121797351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,3584,0.007156444092591603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,5120,0.010798222488827176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,4096,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,3072,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,4096,0.009645333720578088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,3584,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,2560,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,3584,0.009307555854320526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,3072,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,2048,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,3072,0.008405333591832055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,2560,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,1536,0.0042088888585567474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,2560,0.007627555893527136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,2048,0.007548444800906711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,1024,0.0035013332962989807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,2048,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,1536,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,7168,0.012772444221708508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,768,0.0031902222997612427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,1536,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,1024,0.006588444527652528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,512,0.002830222249031067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,7168,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,1024,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,256,0.002471999989615546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,7168,0.013016888664828407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,768,0.005884444548024072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,512,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,256,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,512,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,64,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,256,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,32,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,128,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,128,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,65536,0.035982221364974976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,64,0.005153777698675792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,51200,0.02977688776122199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,65536,0.06343999836179945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,65536,0.0780844423505995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,16384,0.013112000293201871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,51200,0.04974488748444451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,51200,0.061124444007873535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,12288,0.013148444394270578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,16384,0.018823110395007663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,16384,0.023120888405376013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,10240,0.011746666497654386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,768,768,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,12288,0.01566755606068505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,8192,0.010690666735172272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,12288,0.018246221873495314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,10240,0.014444443914625378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,7168,0.009646221995353699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,10240,0.016208888755904306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,8192,0.012344000240166983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,6144,0.009543110926946005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,7168,0.011393778026103973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,7168,0.012800888882742988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,5120,0.008267555799749162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,6144,0.010701333483060202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,768,32,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,4096,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,5120,0.009697777529557547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,5120,0.010703110860453712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,3584,0.00630044440428416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,4096,0.009016888837019602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,4096,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,3072,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,3584,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,3584,0.008979555633332994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,2560,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,128,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,3072,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,3072,0.008273777862389883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,2048,0.004550222307443619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,1536,0.0038648889296584656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,2560,0.007735999921957652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,768,51200,0.043086220820744835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,2560,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,1024,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,2048,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,1536,0.0070257774657673305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,768,0.0028071111689011254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,1536,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,1024,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,1024,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,6144,0.011709333293967776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,256,0.002471111093958219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,768,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,512,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,512,0.0053324442770746015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,128,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,256,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,64,0.0018782222436534034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,256,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,128,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,32,0.0021280000607172647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,65536,0.02424533334043291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,128,0.004970666848950916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,64,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,32,0.005054222212897407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,51200,0.01991555591424306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,65536,0.06026311053170098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,16384,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,65536,0.07698133256700304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,2048,0.007544000115659501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,51200,0.04826044374042087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,512,8192,0.014120888378885059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,12288,0.009696000152164036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,10240,0.008937777744399177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,16384,0.018567111757066514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,16384,0.02202044427394867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,12288,0.015509333875444202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,8192,0.009337777892748514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,10240,0.014450665977266101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,512,768,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,12288,0.017910222212473553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,7168,0.00870666652917862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,6144,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,10240,0.01588800052801768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,8192,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,5120,0.0075786668393347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,8192,0.013762666947311826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,7168,0.011368888947698804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,4096,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,6144,0.010492444038391113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,7168,0.012711111042234631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,3584,0.006260444306664997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,5120,0.009703999592198266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,6144,0.011730666789743634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,4096,0.009015999734401703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,5120,0.010423111418883005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,3072,0.005656888915432825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,3584,0.008626666333940294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,4096,0.009314667019579146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,2560,0.005160000175237656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,3584,0.008815111385451423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,3072,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,2048,0.00452533322903845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,3072,0.008226667013433244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,2560,0.007621333003044128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,1536,0.0038684445122877755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,2560,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,2048,0.007430222299363878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,1024,0.00315022220214208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,2048,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,1536,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,768,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,512,512,0.002583111118939188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,1024,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,512,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,1024,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,768,0.00637155564294921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,256,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,512,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,768,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,128,0.0020711111525694528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,512,0.005335111171007156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,64,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,128,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,256,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,256,32,0.0018871112002266778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,128,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,65536,0.017300445172521803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,64,0.005034666508436203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,51200,0.014309333430396186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,65536,0.05819111400180393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,65536,0.07512178023656209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,16384,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,51200,0.05811466773351034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,51200,0.05957688887914022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,12288,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,1536,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,16384,0.018545778261290658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,16384,0.02219999995496538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,10240,0.008639111287064022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,12288,0.015473778049151102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,8192,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,10240,0.014113777213626437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,12288,0.017877333694034152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,256,256,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,10240,0.015832889411184523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,7168,0.008590222232871586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,6144,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,8192,0.012123555772834353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,7168,0.011384888655609555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,8192,0.013824000126785703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,5120,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,7168,0.012611555556456247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,4096,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,6144,0.011408000356621213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,256,32,0.004957333207130432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,5120,0.009657777845859528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,3584,0.0064533332155810455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,4096,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,5120,0.010337777435779572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,51200,0.046554668082131274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,4096,0.009318222602208456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,3584,0.008642666869693333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,2560,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,3072,0.008150222400824228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,3072,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,2560,0.007719111111429002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,1536,0.0038364442686239877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,2560,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,2048,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,1024,0.0031902222997612427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,2048,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,1536,0.0069857777820693115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,768,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,1536,0.0064275554484791225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,1024,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,512,0.002465777720014254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,1024,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,768,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,256,0.0021564445147911706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,768,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,6144,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,512,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,128,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,512,0.005330666485759947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,256,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,32,0.0019013333237833446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,3072,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,256,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,513,128,128,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,65536,0.015203555425008139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,128,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,64,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,3584,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,16384,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,12288,0.011360888679822287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,65536,0.0749831133418613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,10240,0.009995555712117089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,16384,0.021976888179779053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,8192,0.008719999757077958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,10240,0.015790222419632804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,7168,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,8192,0.01348977784315745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,6144,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,7168,0.01273333364062839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,5120,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,51200,0.013068444199032254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,6144,0.011396444506115384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,4096,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,3584,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,5120,0.010694222317801582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,128,32,0.00519822206762102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,3072,0.005443555613358815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,2560,0.004865777575307422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,4096,0.009337777892748514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,51200,0.05857511361440023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,2048,0.004218666503826777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,3584,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,1536,0.0035386664999855887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,1024,0.003093333293994268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,3072,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,768,0.002568888788421949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,12288,0.01780533293883006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,512,0.0024622221373849446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,2048,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,256,0.002218666589922375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,1536,0.006226666685607698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,128,0.0019022222194406723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,1024,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,64,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,512,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,64,32,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,256,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,128,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,65536,0.0355057782597012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,64,0.004972444640265571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,51200,0.028202666176689997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,16384,0.014813333749771118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,32,0.004859555512666702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,12288,0.020206222931543987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,10240,0.017663111289342243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,128,2048,0.004546666724814309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,65536,0.07419644461737739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,51200,0.05746489100986057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,16384,0.021779555413458083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,12288,0.017588444881969027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,6144,0.011576889289749993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,10240,0.015807999504937064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,8192,0.013472889032628802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,5120,0.010320000350475311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,4096,0.008160889148712158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,2560,0.007400000260935889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,7168,0.012413333687517377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,6144,0.011339555184046427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,3072,0.006724444528420766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,5120,0.010299555957317352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,64,768,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,4096,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,2048,0.004238222208287981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,3584,0.008656889200210571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,1536,0.003531555748648114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,3072,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,1024,0.003156444471743372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,2560,0.007436444362004598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,768,0.0028231110837724474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,2048,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,512,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,1536,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,256,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,1024,0.005761777775155173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,128,0.001863111017478837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,8192,0.014631999863518609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,768,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,64,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,512,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,256,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,128,0.004854222138722737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,64,0.005039111193683413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,3584,0.007535999847782983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,513,32,32,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,2560,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,16384,0.6130693223741319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,16384,0.6677066485087076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,513,32,7168,0.013463111387358772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,16384,0.23631466759575737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,12288,0.47602044211493594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,51200,1.9925102657741969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,51200,0.7153102027045356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,12288,0.1792800029118856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,10240,0.38627465565999347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,51200,2.1929768456353083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,10240,0.15086489253573948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,16384,0.5831199751959907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,10240,0.5000675519307455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,12288,0.4833999739752875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,8192,0.36652178234524196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,8192,0.34316266907585996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,12288,0.4480889108445909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,8192,0.12046578195359971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,7168,0.3286986615922716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,10240,0.3745262093014187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,6144,0.26103732320997447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,7168,0.10664444499545628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,7168,0.2865991062588162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,8192,0.30146222644382054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,5120,0.1890568865670098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,6144,0.09339822000927395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,6144,0.25843466652764213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,4096,0.1541528834236993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,7168,0.27023198869493276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,5120,0.0795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,4096,0.06675999694400363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,4096,0.16921333471934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,6144,0.227347559399075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,5120,0.19092710812886557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,3584,0.14143999417622885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,4096,0.16014311048719618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,3072,0.1128951046201918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,3584,0.059734225273132324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,3584,0.14701333310869005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,2560,0.10891466670566136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,3584,0.13914933469560412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,3072,0.0520648890071445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,2048,0.08682844373914932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,3072,0.12684178352355957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,2560,0.04706311225891113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,3072,0.12202489376068115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,2560,0.11001777648925781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,1536,0.06016533242331611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,2560,0.104312002658844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,2048,0.040349334478378296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,2048,0.09159377548429702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,1024,0.053094221485985645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,2048,0.08830844693713719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,1536,0.07223377625147502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,1536,0.07186222076416016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,768,0.04182666540145874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,1024,0.026613333159022864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,1024,0.05760888920889961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,512,0.028671112325456407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,768,0.02388800018363529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,768,0.04539555642339918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,768,0.04831555485725403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,256,0.018645332919226754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,512,0.021581333544519212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,512,0.03902311126391093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,512,0.040629333919949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,5120,0.21661599477132162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,128,0.01739022301303016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,51200,1.9953689575195312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,256,0.019425777925385367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,256,0.034303112162484065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,128,0.024734222226672705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,64,0.013953778478834363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,128,0.03189155459403992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,65536,32,0.019696000549528334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,128,0.025194666451878015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,64,0.031545778115590416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,32,0.0315688881609175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,256,0.03265955713060167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,1536,0.033470223347345986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,65536,1024,0.05570755402247111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,51200,1.6854498121473525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,65536,0.745338651869032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,65536,2.176347520616319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,65536,2.167670143975152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,16384,0.5357031292385525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,12288,0.40581509802076554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,16384,0.6483555369906956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,51200,0.5683955616421169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,51200,1.7357572979397242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,10240,0.2999937799241808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,12288,0.1424071126514011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,12288,0.38643643591139054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,16384,0.18764711750878227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,16384,0.4617075390285916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,10240,0.3169146643744575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,12288,0.3518346680535211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,8192,0.2845226658715142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,8192,0.0975688894589742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,10240,0.28471289740668404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,7168,0.2239760027991401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,8192,0.23656532499525282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,7168,0.08739022413889568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,6144,0.18265955977969697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,10240,0.11923021740383571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,6144,0.07557155688603719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,8192,0.26075998942057294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,6144,0.19499733712938097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,5120,0.19304354985555014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,6144,0.18010666635301378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,5120,0.16292711098988852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,4096,0.1249119970533583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,5120,0.06450577576955159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,5120,0.15453688303629556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,7168,0.20321688387129042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,51200,1.5206862555609808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,4096,0.054639998409483165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,3584,0.12492088476816814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,4096,0.12472444110446507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,7168,0.23109600279066297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,3584,0.04717688759167989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,3072,0.09373778104782104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,3584,0.11189066039191352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,3584,0.12328267097473145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,2560,0.0976933307117886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,3072,0.04253777861595154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,3072,0.10155022144317627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,3072,0.097561776638031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,2048,0.06417422162161933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,2560,0.037929776642057635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,2560,0.09007289012273152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,1536,0.04940088921123081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,2048,0.033039109574423894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,2048,0.07342222001817492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,4096,0.13209333684709337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,1024,0.035750223530663386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,1536,0.058915555477142334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,1536,0.02747022277779049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,65536,2.058902316623264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,1536,0.05918577644560072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,1024,0.02053955528471205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,2048,0.0718382199605306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,512,0.024328889118300542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,1024,0.04778666628731621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,768,0.018369777335060965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,768,0.036710222562154136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,256,0.0181404451529185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,512,0.017064889272054035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,512,0.03035377793841892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,512,0.034770667552948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,256,0.015549333559142219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,2560,0.08413244618309869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,256,0.02685600022474925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,256,0.027653333213594224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,64,0.014212444424629211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,51200,128,0.015669332610236276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,128,0.02574399941497379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,32,0.014736889137162102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,128,0.021519111262427434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,64,0.02567377852069007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,768,0.035713778601752386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,32,0.025285333395004272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,51200,1024,0.044962667756610446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,65536,0.5944026841057671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,51200,768,0.041020443042119346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,51200,0.459030204349094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,65536,0.258462217119005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,65536,0.7039884461296929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,16384,0.15454310841030544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,51200,0.5457644462585449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,51200,0.20027821593814424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,16384,0.06714222166273329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,16384,0.18405777878231475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,12288,0.12124000655280219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,10240,0.0964755548371209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,16384,0.16559022002749973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,12288,0.12821333938174778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,12288,0.05241866575347053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,10240,0.04396533303790622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,12288,0.12768888473510742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,10240,0.10794222354888916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,8192,0.10026222467422485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,10240,0.10892089207967122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,7168,0.0725875563091702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,51200,128,0.015831111205948722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,8192,0.036729776197009616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,8192,0.08819733063379924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,6144,0.06951822174919976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,8192,0.08988355265723334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,7168,0.0330079992612203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,7168,0.07793600029415555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,5120,0.053492443429099194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,65536,0.6464728779262966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,7168,0.08066311147477892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,51200,0.5028844409518772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,4096,0.049513777097066246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,6144,0.028743998871909246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,6144,0.07089689042833117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,5120,0.028392000330819026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,5120,0.058336887094709605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,3584,0.043868445687823825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,4096,0.021516443954573736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,5120,0.06291911337110731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,3072,0.038768887519836426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,3584,0.01943822205066681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,4096,0.0524088872803582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,3584,0.042993777328067355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,2560,0.031425777408811785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,3584,0.047725333107842334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,3072,0.017672888106769986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,3072,0.0377493335141076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,3072,0.043142222695880465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,2048,0.025666667355431452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,2560,0.016415110892719693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,2560,0.03300088975164626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,2560,0.03779911001523336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,1536,0.0199644449684355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,2048,0.014677332507239448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,2048,0.0273982220225864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,1024,0.014465777410401238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,1536,0.012707555459605323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,1536,0.02256177696916792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,6144,0.06881511211395264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,768,0.0123857781291008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,1536,0.02809600035349528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,1024,0.0107342220014996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,4096,0.04825955629348755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,1024,0.023246222072177466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,768,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,512,0.008988444175985124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,768,0.020611556039916146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,256,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,512,0.009338666995366415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,512,0.013088000317414602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,512,0.018391110830836825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,128,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,256,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,256,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,256,0.015834665960735746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,64,0.005139555368158552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,128,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,16384,32,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,128,0.011360888679822287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,2048,0.03348177671432495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,128,0.012107555237081317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,64,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,32,0.011406222151385413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,1024,0.01799199978510539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,16384,768,0.015171556009186639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,65536,0.48337422476874453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,51200,0.36419910854763454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,65536,0.22135289510091147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,65536,0.5376186900668675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,16384,0.11837955315907796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,51200,0.4230968952178955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,51200,0.16658488909403482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,16384,0.05491822295718723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,12288,0.09051199754079182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,16384,0.13342932860056558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,16384,0.13016977575090197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,12288,0.04200977749294705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,12288,0.10179911057154338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,12288,0.10074933369954427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,10240,0.0857342216703627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,10240,0.034770667552948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,8192,0.08476799726486206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,7168,0.07492978043026395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,10240,0.08765777614381577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,8192,0.028746667835447524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,65536,0.4910435676574707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,8192,0.07001777489980061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,51200,0.3906551202138265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,6144,0.05995555718739828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,8192,0.0727528863483005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,7168,0.062111112806532115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,5120,0.046203556987974376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,6144,0.02474044428931342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,6144,0.054930667082468666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,4096,0.041858666472964816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,5120,0.020933333370420668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,5120,0.046760890218946666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,5120,0.05087200138303968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,10240,0.08624177508884007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,4096,0.0181342214345932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,3584,0.03303822212749057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,4096,0.03879911038610671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,3072,0.031181332137849595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,4096,0.04371822211477491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,3584,0.03417244553565979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,3584,0.03938400083118015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,3072,0.030768000417285498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,3072,0.015219555960761176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,3072,0.03545422355333964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,7168,0.02591466738118066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,2048,0.02106133268939124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,2560,0.02668711046377818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,2560,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,7168,0.065829336643219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,1536,0.01571822166442871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,2560,0.03127822279930115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,2048,0.012702222499582501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,6144,0.060193777084350586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,2048,0.022742221752802532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,1024,0.012753778033786349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,2048,0.02774488925933838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,1536,0.010919111470381418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,1536,0.01891377733813392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,768,0.010647111468844943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,1536,0.024097777075237695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,1024,0.009334222310119206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,1024,0.015119110544522604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,512,0.00832177781396442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,3584,0.016719111137919955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,768,0.008655110994974772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,768,0.012776888906955719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,768,0.017966222431924608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,256,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,2560,0.025962667332755193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,512,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,512,0.011032000184059143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,128,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,512,0.01702044407526652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,256,0.007928888830873702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,256,0.010019555687904358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,64,0.005497777627574072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,12288,32,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,128,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,128,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,128,0.011495999991893768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,64,0.009753777749008602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,12288,32,0.009658666948477427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,65536,0.40216888321770566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,51200,0.3325084315405952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,65536,0.4452844195895725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,1024,0.020099555452664692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,16384,0.11319200197855632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,51200,0.35899199379815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,51200,0.14157866107092962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,256,0.014370666609870063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,12288,0.08190844456354777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,16384,0.11260355843438043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,16384,0.04865600003136528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,10240,0.06524888674418132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,12288,0.08557599782943726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,16384,0.12725867165459528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,65536,0.4879804187350803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,12288,0.09871021906534831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,8192,0.06466577450434367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,10240,0.07205688953399658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,65536,0.17516622278425428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,51200,0.38354044490390354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,10240,0.08580355511771308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,8192,0.027112889620992873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,8192,0.059788445631663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,6144,0.03942044576009115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,8192,0.07214133607016669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,7168,0.024640889631377325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,7168,0.053175111611684166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,5120,0.04018933243221707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,6144,0.021969777014520433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,6144,0.046688887808057994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,6144,0.05633955531650119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,5120,0.019632889164818656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,4096,0.03236711025238037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,5120,0.040044443474875555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,5120,0.04943199952443441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,4096,0.017243555850452848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,3584,0.028453333510292902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,4096,0.03308799862861633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,10240,0.03306577934159173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,3072,0.02456088860829671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,3584,0.01605422298113505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,3584,0.02982577681541443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,7168,0.05422311027844747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,3584,0.038429333104027644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,2560,0.01956533392270406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,3072,0.014783999986118741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,3072,0.03474311033884684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,2560,0.013639111485746173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,2560,0.02313511073589325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,2048,0.01572088897228241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,7168,0.06395022074381511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,2560,0.03066666589842902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,1536,0.01310400002532535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,2048,0.012094222009181976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,12288,0.03882666760020786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,2048,0.0199608885579639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,2048,0.02698844505680932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,1536,0.010403555300500657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,1536,0.02346933384736379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,768,0.008385777473449707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,1024,0.008958222137557136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,1024,0.013350221845838757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,4096,0.04182844360669454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,512,0.006691555596060223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,768,0.008348444269763099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,768,0.01697688963678148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,3072,0.026698667142126296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,256,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,512,0.008039111064540016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,512,0.01512888901763492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,256,0.009387555221716562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,128,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,256,0.013179555535316467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,1024,0.009859555297427708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,1536,0.016781333420011733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,64,0.004342222379313575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,128,0.009004444711738164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,10240,32,0.004559111264016894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,128,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,128,0.010840000377760993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,1024,0.019499555230140686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,64,0.008983111215962304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,768,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,32,0.008987555901209513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,65536,0.2551182111104329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,10240,512,0.009739555418491364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,51200,0.20029777950710723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,65536,0.13429066869947645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,65536,0.3659537898169623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,256,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,16384,0.09283644623226589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,51200,0.2968222300211589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,51200,0.10708355241351658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,12288,0.05376977721850077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,16384,0.03806311223242018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,16384,0.09200088845358954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,10240,0.04428622126579285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,16384,0.0940542221069336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,12288,0.029746668206320867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,12288,0.07024177577760485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,12288,0.07402577665117052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,8192,0.047784888082080416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,65536,0.3401830991109212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,51200,0.2699893315633138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,10240,0.025843555728594463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,7168,0.04227111074659559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,10240,0.05958933300442166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,8192,0.021328000558747187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,10240,0.0639244450463189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,6144,0.03552000059021844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,8192,0.049364444282319814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,7168,0.019348444210158456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,7168,0.04407822092374166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,5120,0.030785779158274334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,6144,0.01793244481086731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,6144,0.038486222426096596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,4096,0.024783111280865137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,6144,0.04407911168204414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,5120,0.033406221204333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,5120,0.03840444485346476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,3584,0.02162577708562215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,4096,0.0277839998404185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,4096,0.03300088975164626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,3584,0.024893333514531452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,3584,0.01330488920211792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,3584,0.030100444952646892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,3072,0.02239022155602773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,3072,0.01221777747074763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,2560,0.016114667057991028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,8192,0.055120001236597695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,3072,0.027408889598316614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,2048,0.013769778112570444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,7168,0.0495742228296068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,2560,0.011232000258233813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,2560,0.019727110862731934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,1536,0.011159111228254108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,2560,0.024545777175161574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,2048,0.009869332942697737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,2048,0.017144888639450073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,1024,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,2048,0.021928000781271193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,1536,0.008637333081828224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,1536,0.014429334137174817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,768,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,1536,0.019433778193261888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,1024,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,1024,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,3072,0.01865333318710327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,1024,0.016185777054892648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,768,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,768,0.01000711073478063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,256,0.004520888957712385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,768,0.01477777792347802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,512,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,512,0.008760888543393876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,128,0.003871110992299186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,512,0.013167111410035027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,256,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,256,0.012200000385443369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,128,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,128,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,32,0.0038559999730851916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,5120,0.01593777702914344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,128,0.009749333063761393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,4096,0.014387556248241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,64,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,32,0.008374222450786168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,65536,0.30249423450893825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,512,0.0058471109304163195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,51200,0.2374337779151069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,65536,0.1281191110610962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,16384,0.09004444546169704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,51200,0.2679351170857747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,51200,0.10139111015531753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,8192,256,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,12288,0.0681519971953498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,16384,0.08764444457160102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,8192,64,0.0035573335157500375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,16384,0.09325777822070652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,51200,0.2634630997975667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,12288,0.06730844577153523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,10240,0.05498488744099935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,12288,0.028935111231274072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,10240,0.024914667010307312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,10240,0.05728977918624878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,8192,0.04699022240108914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,65536,0.34218133820427793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,10240,0.0643635524643792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,7168,0.039439111948013306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,8192,0.02110222147570716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,8192,0.0473217765490214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,8192,0.0532186660501692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,6144,0.035327110025617815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,7168,0.04222222169240316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,16384,0.03710844450526767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,7168,0.04798400070932177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,5120,0.02936533424589369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,65536,0.33863200081719297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,6144,0.017807111144065857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,6144,0.03710666630003188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,4096,0.024244444237815008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,6144,0.04337333308325874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,5120,0.015998222761683993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,5120,0.031932443380355835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,3584,0.02163733376397027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,5120,0.03804444604449802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,4096,0.026965333355797663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,4096,0.03242044316397773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,3072,0.019466666711701285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,3584,0.013109332985348172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,3584,0.024438222249348957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,3584,0.02985244327121311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,3072,0.012069332930776807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,3072,0.02197333342499203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,7168,0.019338667392730713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,3072,0.026923555466863845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,2048,0.014182221558358936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,2560,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,2560,0.019268444842762418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,1536,0.011928888658682505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,2560,0.02444533341460758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,2048,0.016508445143699646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,2048,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,12288,0.07447022199630737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,2048,0.02197511163022783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,1536,0.008605333666006723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,1536,0.013977777626779346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,4096,0.014365333649847241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,768,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,1536,0.019578667150603402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,1024,0.007931555310885111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,1024,0.011009777585665384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,512,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,768,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,1024,0.01614755557643043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,768,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,256,0.004997333304749595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,768,0.014370666609870063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,512,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,128,0.0042204442951414315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,2560,0.016855110724767048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,512,0.01314755529165268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,256,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,64,0.003908444609906939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,256,0.008029333419269985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,128,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,256,0.012509333590666452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,128,0.007710222568776872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,128,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,64,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,32,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,1024,0.00941066692272822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,51200,0.19462845060560438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,65536,0.2982470989227295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,51200,0.23281600740220812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,51200,0.09717333316802979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,7168,512,0.008635555704434713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,65536,0.33809600936041934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,7168,32,0.005136888888147143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,16384,0.07761422130796644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,12288,0.0510808891720242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,51200,0.26571912235683864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,16384,0.03564000129699707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,16384,0.09405866596433853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,10240,0.04383111000061035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,65536,0.25233689943949383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,12288,0.027829332484139338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,12288,0.05995821952819824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,8192,0.040115555127461754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,12288,0.0730453332265218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,10240,0.05090488990147909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,10240,0.02433066732353634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,65536,0.12150843938191731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,7168,0.03426311082310147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,10240,0.0635475582546658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,8192,0.020584000481499564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,8192,0.04199111130502489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,6144,0.028704888290829126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,7168,0.018968888454967074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,8192,0.053156445423762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,7168,0.037589334779315524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,5120,0.02494311167134179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,16384,0.07723822196324666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,7168,0.04753066764937507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,6144,0.017323555217848886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,4096,0.02015466656949785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,5120,0.015625778171751235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,6144,0.04284177886115181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,5120,0.029064890411165025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,3584,0.01724088854259915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,5120,0.03723377651638455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,4096,0.013776000175211163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,3072,0.01633777717749278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,4096,0.03204177816708883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,3584,0.012704888979593912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,3584,0.02221511138810052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,2560,0.014128888646761576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,3584,0.029513776302337646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,3072,0.011727111207114326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,3072,0.0199635558658176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,2048,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,3072,0.026882666680547927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,2560,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,2560,0.01757866640885671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,1536,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,2048,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,2560,0.024005333582560223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,2048,0.015098666151364645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,1024,0.007729777859316931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,2048,0.021867555048730638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,1536,0.00832177781396442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,1536,0.012416000167528788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,6144,0.03362400001949734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,768,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,1536,0.01891911029815674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,1024,0.007709333466158972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,1024,0.010394666757848527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,1024,0.015578667322794596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,768,0.007592889169851939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,4096,0.024439111351966858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,768,0.009353777600659264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,256,0.004509333521127701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,768,0.014544000228246054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,512,0.007234666910436418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,512,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,128,0.0038942222793896994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,256,0.00685422205262714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,512,0.013432888521088494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,256,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,64,0.003605333467324575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,32,0.0038728887836138406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,128,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,128,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,128,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,64,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,65536,0.18694399462805855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,6144,32,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,51200,0.1382311052746243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,65536,0.2566995620727539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,16384,0.06176177660624186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,6144,512,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,51200,0.19900088840060762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,51200,0.09519377681944106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,12288,0.04676533407635159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,16384,0.033839109871122576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,16384,0.06739377975463867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,51200,0.26411377059088814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,65536,0.33766844537523055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,10240,0.03172355559137132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,256,0.011878222227096558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,12288,0.05151999990145365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,8192,0.031657778554492526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,12288,0.07219555642869738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,10240,0.023774221539497375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,10240,0.04438044296370613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,7168,0.02784622377819485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,10240,0.06213777595096164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,8192,0.019952888290087383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,8192,0.03644444545110067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,6144,0.023973332511054143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,8192,0.052375111314985484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,7168,0.018401778406567044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,7168,0.03261244297027588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,5120,0.020375111036830477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,7168,0.046690666013293795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,6144,0.029098666376537744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,6144,0.041466666592492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,5120,0.025395555628670588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,5120,0.015211555692884656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,4096,0.01645066671901279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,16384,0.0923715564939711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,5120,0.03673244516054789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,3584,0.014560888210932413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,12288,0.02682311170630985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,4096,0.02164622147878011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,3072,0.01275644451379776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,4096,0.031298667192459106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,3584,0.012076444096035428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,3584,0.019501333435376484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,3584,0.028765334023369685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,2560,0.01128888875246048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,3072,0.017841777867741056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,3072,0.011086222198274402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,3072,0.026340444882710774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,2048,0.009650666680600908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,65536,0.11928799417283799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,2560,0.009932444327407414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,2560,0.015477332803938123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,1536,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,6144,0.016691555579503376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,2560,0.023580445183648005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,2048,0.008961777720186446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,2048,0.013421333498424955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,1024,0.00646311127477222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,2048,0.021310221817758348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,1536,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,768,0.005524444497293896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,1536,0.01829955478509267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,1024,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,4096,0.013125333521101209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,1024,0.009324444664849175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,512,0.00452888881166776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,1024,0.015216888652907478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,768,0.007361777954631382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,256,0.003546666767862108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,768,0.014110222458839417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,512,0.007601777712504069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,512,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,128,0.003208000005947219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,256,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,256,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,256,0.011944889194435544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,128,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,128,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,32,0.0032097777972618737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,128,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,64,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,65536,0.1719591087765164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,32,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,1536,0.011119999819331698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,51200,0.1424471139907837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,65536,0.07608799801932441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,65536,0.21044711271921793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,512,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,16384,0.04609866605864631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,51200,0.1660648849275377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,51200,0.061408003171284996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,65536,0.1968933343887329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,12288,0.03996088769700792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,16384,0.05650844176610311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,16384,0.02271644439962175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,51200,0.15502133634355333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,5120,64,0.003157333367400699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,10240,0.03293155630429586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,16384,0.058761775493621826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,12288,0.018579555882347953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,12288,0.04360533422893948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,8192,0.024977778395016987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,10240,0.016567111015319824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,12288,0.04691733254326714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,7168,0.023153778579499986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,10240,0.037288000186284385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,8192,0.014262222581439547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,10240,0.04128266705407037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,6144,0.019328888919618394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,7168,0.013287111288971372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,8192,0.03494844502872891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,5120,0.016433777080641854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,7168,0.032283554474512734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,6144,0.01221422188811832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,6144,0.02514311174551646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,4096,0.01440088947614034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,6144,0.028767112228605483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,5120,0.010983111129866706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,5120,0.022572444544898138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,3584,0.012576888832781049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,5120,0.025588444537586633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,4096,0.009639110830095079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,4096,0.018940443793932598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,3072,0.011681777735551199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,3584,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,3584,0.017236444685194228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,3584,0.021320889393488567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,3072,0.015626667274369132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,3072,0.008678221868144141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,3072,0.019592000378502738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,8192,0.03113244308365716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,2048,0.008758222063382467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,2560,0.008276444342401292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,7168,0.028312000963422988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,2560,0.01776000029510922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,5120,768,0.008344888687133789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,2048,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,2048,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,2048,0.015663110547595553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,1024,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,1536,0.00999111102686988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,1536,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,1536,0.014372444815105863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,4096,0.02233155568440755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,1024,0.008627555436558193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,768,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,1024,0.01240799989965227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,2560,0.010017777482668558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,768,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,768,0.007669332954618666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,768,0.011517333487669626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,2560,0.013250666360060373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,512,0.007037333316273159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,512,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,1536,0.006743999818960826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,512,0.011411555939250521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,128,0.002946666752298673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,256,0.006230222268237009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,256,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,256,0.01034400032626258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,64,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,128,0.006333333336644703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,128,0.006223111102978389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,32,0.003145777723855443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,128,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,65536,0.14781867133246526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,1024,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,64,0.006593777901596493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,4096,32,0.006567111031876669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,512,0.0038968887594011095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,65536,0.07328177822960748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,16384,0.048103110657797925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,65536,0.2077982160780165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,51200,0.058746669027540416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,51200,0.16260977586110434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,4096,256,0.003488000068399641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,12288,0.034098668230904475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,65536,0.19288710753122965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,16384,0.022264000442292955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,16384,0.05505599909358554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,10240,0.029176000091764662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,12288,0.018163555198245578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,16384,0.05752444267272949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,12288,0.04239555531077915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,8192,0.02253688871860504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,12288,0.045575112104415894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,10240,0.016396444704797532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,10240,0.03644355469279819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,7168,0.020984888076782227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,10240,0.040476444694730974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,8192,0.03046400017208523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,6144,0.017995556195576985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,8192,0.03418755531311035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,7168,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,7168,0.027426666683620874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,5120,0.015266666809717814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,7168,0.03179377648565505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,51200,0.11650489436255561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,6144,0.011874666644467248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,6144,0.02453600035773383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,4096,0.01314399970902337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,5120,0.010501333408885531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,6144,0.028395556741290625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,5120,0.021556443638271753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,3584,0.011406222151385413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,5120,0.025519111090236243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,4096,0.01850488947497474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,51200,0.15253067016601562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,3072,0.010814222196737925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,3584,0.009273777405420939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,4096,0.021735111872355144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,3584,0.016564443707466125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,2560,0.009332444104883406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,3584,0.020609777834680345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,3072,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,3072,0.014633778068754407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,3072,0.019352888067563374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,2560,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,8192,0.014087110757827759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,2560,0.017208000024159748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,2048,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,2048,0.011143999795118967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,2048,0.015580443872345818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,1024,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,1536,0.007328888608349695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,1536,0.009735110733244155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,1536,0.013834666874673633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,768,0.004853333450025982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,1024,0.006783999916579988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,1024,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,1024,0.012437333663304647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,4096,0.009371555513805812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,512,0.004226666771703296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,768,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,768,0.007620444728268518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,256,0.0035333335399627686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,768,0.011730666789743634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,512,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,512,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,2048,0.008242666721343994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,128,0.0031600000543726813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,512,0.010814222196737925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,256,0.006009777800904379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,2560,0.01276000009642707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,256,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,64,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,32,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,128,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,128,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,128,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,65536,0.14872977468702528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,32,0.006606222026877933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,51200,0.11150044865078396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,65536,0.0716444452603658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,65536,0.17103021674686006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,65536,0.18870133823818633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,51200,0.05749155415429009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,16384,0.038365334272384644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,51200,0.1331119934717814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,51200,0.1506684488720364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,16384,0.021574222379260596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,16384,0.04577777783075968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3584,1536,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,16384,0.057384888331095375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,10240,0.02477600011560652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,12288,0.03539377782079909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,12288,0.01795111099878947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3584,64,0.006462222172154321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,8192,0.021222222182485793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,12288,0.04544088906712002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,10240,0.01622577839427524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,7168,0.019141332970725167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,10240,0.030811554855770532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,8192,0.01385155568520228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,10240,0.03977955712212457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,8192,0.025975111458036635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,6144,0.01628977722591824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,8192,0.03397511111365424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,7168,0.012744888663291931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,5120,0.014184888866212634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,6144,0.011744889120260874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,7168,0.030994666947258845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,6144,0.021191999316215515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,4096,0.01205599970287747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,6144,0.02812533246146308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,5120,0.018889778190188937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,5120,0.010553778045707278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,12288,0.031938665442996554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,3584,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,5120,0.025376889440748427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,4096,0.009328888522254096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,256,0.010709332923094431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,4096,0.01590133375591702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,3072,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,4096,0.02202044427394867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,3584,0.014472888575659858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,2560,0.008563555777072906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,3584,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,3072,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,3584,0.0196435550848643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,2048,0.007465778125656976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,2560,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,3072,0.012763555679056378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,1536,0.006324444380071428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,2560,0.01127199994193183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,2048,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,2048,0.009895111123720804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,1024,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,2048,0.015166223049163818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,1536,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,1536,0.008843555218643611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,7168,0.023409777217441138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,768,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,1536,0.013915555344687568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,1024,0.006722666737106111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,512,0.0038071109188927543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,1024,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,768,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,1024,0.012740444805887012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,768,0.0069146665434042616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,768,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,512,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,512,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,128,0.0031662223239739737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,512,0.010821333361996545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,256,0.005958222266700532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,256,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,64,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,256,0.01072177787621816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,128,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,32,0.002903999967707528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,128,0.005962666538026597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,3072,0.018568888306617737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,128,0.008088000118732452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,65536,0.13091199927859837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,64,0.005967111223273807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,2560,0.01668444441424476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,3072,32,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,51200,0.09908711247973973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,16384,0.037321776151657104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,65536,0.06949244605170356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,65536,0.1650008890363905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,51200,0.056306666798061795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,51200,0.1296693351533678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,65536,0.19023911158243814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,12288,0.029142220815022785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,51200,0.15170578161875406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,16384,0.021264889174037512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,16384,0.04459911253717211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,3072,256,0.003185777821474605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,10240,0.023804444405767653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,12288,0.017794667018784415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,12288,0.0347671111424764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,10240,0.015846222639083862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,12288,0.04509600003560384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,10240,0.0303422212600708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,7168,0.017065778374671936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,10240,0.03941599859131707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,8192,0.013390222357379066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,8192,0.025424000289705064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,8192,0.03348088926739163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,6144,0.014913777510325113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,7168,0.01238666640387641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,7168,0.022989332675933838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,5120,0.013357333011097379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,7168,0.031173331869973078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,6144,0.011222222612963783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,6144,0.020809777908855014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,4096,0.011231111155615913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,6144,0.02776977750990126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,5120,0.009983110758993361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,5120,0.01799199978510539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,3584,0.01016444464524587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,5120,0.024995555480321247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,4096,0.00925511121749878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,4096,0.014959999256663851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,16384,0.056399113602108426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,3072,0.009177777502271865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,3584,0.008645333349704742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,3584,0.013796444568369122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,8192,0.018481777773963082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,3584,0.020637333393096924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,3072,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,3072,0.012438221938080259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,2048,0.0069146665434042616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,3072,0.018233777748213876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,2560,0.007818666597207388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,2560,0.011019555230935415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,1536,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,2560,0.01682400041156345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,2048,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,2048,0.014854222536087036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,1024,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,1536,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,1536,0.014150222142537435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,768,0.0038746665749284956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,1024,0.0075662218862109715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,1024,0.006727111008432176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,1024,0.012706666356987424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,512,0.003395555540919304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,4096,0.021018667353524104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,768,0.0063902222447925145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,768,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,256,0.0029191111938820947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,2560,0.008303110798199972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,512,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,512,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,128,0.0027964444210131965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,512,0.010422222316265106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,256,0.006083555519580841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,256,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,2048,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,128,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,128,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,32,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,128,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,65536,0.09288533528645833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,1536,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,64,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2560,32,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,51200,0.07993244462543063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,65536,0.06838755475150214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,65536,0.12861955165863037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,16384,0.034197333786222674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,51200,0.05520977907710605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,51200,0.10071200132369995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,768,0.011373333632946014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,51200,0.14345955848693848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,12288,0.025847999585999384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,16384,0.020851555797788832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,16384,0.03494844502872891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,256,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,12288,0.017208000024159748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,12288,0.027657777070999146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2560,64,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,8192,0.018043556147151522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,12288,0.04315822323163351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,10240,0.024466666910383437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,10240,0.015224888920783997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,7168,0.016015999846988253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,10240,0.037343998750050865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,8192,0.012531555361217923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,8192,0.020650666620996263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,8192,0.0321999986966451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,7168,0.01889866590499878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,7168,0.011834666960769229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,7168,0.029764443635940552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,65536,0.18094755543602836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,5120,0.012110222544935016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,6144,0.016727111405796476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,6144,0.01072355525361167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,6144,0.026913776993751526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,4096,0.010473777850468954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,5120,0.010085333552625444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,10240,0.021964444054497614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,5120,0.014961777461899651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,16384,0.054543998506334096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,3584,0.009395555489593083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,5120,0.023189332750108507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,4096,0.009054222040706212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,3072,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,4096,0.012525333298577202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,4096,0.020051555501090158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,3584,0.01148088855875863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,3584,0.018939556346999276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,3072,0.008275555239783393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,3072,0.010568888651000129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,3072,0.01724177764521705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,2048,0.0063288890653186375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,6144,0.014056000444624158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,2560,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,2560,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,1536,0.005461333526505365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,2560,0.015839111473825242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,2048,0.0076657773719893555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,1024,0.004450666821665234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,2048,0.00867022242810991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,1536,0.0070293330483966405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,2048,0.01460977726512485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,1536,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,768,0.003913777983850903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,1536,0.013278221918476952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,1024,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,1024,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,512,0.0031644445326593188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,1024,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,768,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,768,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,256,0.0031884445084465873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,768,0.011384000380833944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,512,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,3584,0.00867911097076204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,512,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,512,0.011009777585665384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,256,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,2560,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,64,0.0026542221506436667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,256,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,32,0.0028320000403457214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,256,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,128,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,128,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,65536,0.0889973309304979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,128,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,64,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,51200,0.06844355662663777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,2048,32,0.005678222411208683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,65536,0.06701511144638062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,16384,0.026586666703224182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,65536,0.12166133191850449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,51200,0.054027554061677724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,12288,0.019280888968043856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,51200,0.0953777763578627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,16384,0.02077155477470822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,16384,0.03392088744375441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,51200,0.1423599984910753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,10240,0.016526222229003906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,16384,0.05282310975922478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,12288,0.016571554872724745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,12288,0.02725511127048069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,10240,0.015012444721327888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,12288,0.04191111193762886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,10240,0.02368444369898902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,2048,128,0.0028479999552170434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,10240,0.036848889456854925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,8192,0.019833778341611225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,8192,0.012497777740160624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,6144,0.011056000159846412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,7168,0.011423110961914062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,7168,0.017887999614079792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,7168,0.0284222231970893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,5120,0.009854222337404886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,6144,0.01589866644806332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,6144,0.0107004443804423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,6144,0.02548177705870734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,5120,0.009697777529557547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,4096,0.008650666309727563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,5120,0.014467555615637036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,5120,0.022871999277008906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,3584,0.007687999970383114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,4096,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,4096,0.012243555651770698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,8192,0.013824889229403602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,4096,0.02000711030430264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,3072,0.006984000404675801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,3584,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,3584,0.008642666869693333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,7168,0.01200355589389801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,3584,0.018392889036072623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,3072,0.00832799987660514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,3072,0.01034577770365609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,8192,0.031898667414983116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,3072,0.017027555240525138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,2560,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,2560,0.009308444129096137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,2560,0.015821332732836407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,2048,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,1536,0.004553777890072929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,2048,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,2048,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,65536,0.17938844362894693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,1536,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,1536,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,1536,0.012837332983811697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,768,0.0035146665241983202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,1024,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,1024,0.0068604445291890045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,1024,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,512,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,768,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,2560,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,768,0.011032000184059143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,256,0.0027928888383838865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,512,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,512,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,2048,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,128,0.002467555511328909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,512,0.01036977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,256,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,256,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,64,0.002459555657373534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,256,0.009659555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,128,0.005444444302055571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,32,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1536,1024,0.0039013334446483185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,65536,0.05089866783883837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,64,0.005394666559166378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1536,32,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,51200,0.04378488990995619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,65536,0.06569688849978976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,768,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,16384,0.01754133403301239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,51200,0.052480889691246875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,51200,0.06366044282913208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,51200,0.1405786673227946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,12288,0.017606221967273288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,16384,0.02401955591307746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,16384,0.05155200097295973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,12288,0.015855110353893705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,12288,0.01926399932967292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,10240,0.01555022266176012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,128,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,12288,0.04145510991414388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,8192,0.013003555436929068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,10240,0.013939556148317126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,10240,0.016846223009957206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,65536,0.08060089084837171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,10240,0.035996443695492215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,7168,0.012010667059156628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,8192,0.012114666402339935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,8192,0.031050665511025324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,65536,0.17696355448828804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,6144,0.010625777973069085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,7168,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,7168,0.013210666676362356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,5120,0.009292444421185387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,7168,0.02812088860405816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,6144,0.010392888552612728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,16384,0.01975555552376641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,6144,0.01202666676706738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,4096,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,6144,0.024939555260870192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,5120,0.01090222183201048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,3584,0.007231999602582719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,5120,0.02314577831162347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,128,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,4096,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,4096,0.009675555759006077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,3072,0.006567111031876669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,4096,0.01945600079165565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,3584,0.008358221915033128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,3584,0.00923999978436364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,2560,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,3072,0.00814400033818351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,3072,0.008610666626029545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,3584,0.017902221944597032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,2048,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,3072,0.016907556189431083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,8192,0.014490667316648694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,1536,0.004247111164861255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,2560,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,2560,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,2560,0.015485333071814643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,2048,0.0070568886068132185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,1024,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,2048,0.013912000589900546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,1536,0.007024889190991719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,1536,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,768,0.003149333306484752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,1536,0.013409777647919126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,1024,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,1024,0.006031999985376994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,5120,0.009668444593747457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,512,0.0028168888141711554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,1024,0.011700444751315646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,768,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,768,0.005656888915432825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,256,0.002493333278430833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,768,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,512,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,512,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,256,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,512,0.010317333042621613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,256,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,64,0.0021217777911159727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,256,0.00941955546538035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,32,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,128,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,128,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,65536,0.04389422138532003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,128,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,64,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,2048,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,51200,0.036080890231662326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,1024,32,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,65536,0.06390755706363253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,65536,0.07947733004887898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,16384,0.015504888362354703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,51200,0.05063555637995402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,51200,0.06251644425921969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,65536,0.17671733432345918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,12288,0.014125333891974555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,16384,0.01886666648917728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,51200,0.14059378041161433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,16384,0.023416888382699754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,10240,0.01278222186697854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,16384,0.05128533310360379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,12288,0.01533422205183241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,8192,0.011943110989199745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,10240,0.01660977800687154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,10240,0.01408088869518704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,10240,0.035306665632459856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,7168,0.011726222104496427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,8192,0.01202666676706738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,8192,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,8192,0.030050665140151978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,6144,0.010310221877362993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,7168,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,7168,0.02754666739039951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,5120,0.008918222453859117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,6144,0.012324444121784635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,6144,0.01035199976629681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,6144,0.024791111548741657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,4096,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,5120,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,5120,0.010673777924643623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,12288,0.01868711080816057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,4096,0.009015999734401703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,3584,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,4096,0.009696000152164036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,4096,0.019183999962276883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,3584,0.008608000146018134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,3072,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,3584,0.009005332986513773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,3584,0.017906667457686532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,2560,0.005777777896987067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,3072,0.008083555433485243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,7168,0.01110400011142095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,3072,0.00851999968290329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,2048,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,2560,0.007687110867765214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,2560,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,2560,0.01517688896920946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,2048,0.007080000307824876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,1536,0.0042408886882993905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,2048,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,2048,0.013821333646774292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,1536,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,1024,0.003543111185232798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,5120,0.022080888350804646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,1536,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,1536,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,1024,128,0.0021911110315057966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,1024,0.006588444527652528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,1024,0.005962666538026597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,512,0.0028773333049482773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,1024,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,768,0.005723555468850666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,768,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,256,0.002550222186578645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,768,0.010845333337783813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,512,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,512,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,3072,0.01686311099264357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,128,0.0026951111439201566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,256,0.005384888913896348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,256,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,256,0.009324444664849175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,64,0.002199999988079071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,128,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,32,0.002492444382773505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,128,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,128,0.007739555504586961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,65536,0.03183999988767836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,64,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,768,32,0.005173333403136996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,768,768,0.0032142222755485107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,12288,0.04068444503678216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,51200,0.02659911082850562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,65536,0.06390221913655598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,65536,0.07806399795744154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,16384,0.013384000294738345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,65536,0.17714756064944795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,51200,0.06195911433961657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,12288,0.01128088848458396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,16384,0.022966222630606756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,51200,0.13984978199005127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,16384,0.051078223519855075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,10240,0.011614222493436603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,512,0.010028444230556488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,12288,0.015288889408111572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,12288,0.01846933364868164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,8192,0.012126222252845764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,12288,0.04091822107632955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,10240,0.016185777054892648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,7168,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,8192,0.014134221606784396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,8192,0.011775111158688864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,8192,0.029726223813162908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,6144,0.009527111219035255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,7168,0.011077333655622272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,7168,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,7168,0.02716444432735443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,5120,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,51200,0.05032799972428215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,6144,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,6144,0.012104888757069906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,4096,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,6144,0.024041778511471216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,16384,0.018238221605618794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,5120,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,3584,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,5120,0.010788444843557147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,5120,0.02168088820245531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,4096,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,4096,0.009654222263230218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,3072,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,4096,0.018995554910765756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,3584,0.008981333010726506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,10240,0.013818666338920593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,2560,0.005186666631036335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,3584,0.017866666118303936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,3072,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,3072,0.00830844458606508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,2048,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,3072,0.016571554872724745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,2560,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,2560,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,1536,0.0038728887836138406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,2560,0.015223110715548197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,2048,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,1024,0.0032542221662071017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,2048,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,2048,0.013763555222087435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,1536,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,768,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,1536,0.012740444805887012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,1024,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,1024,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,512,0.0024871110088295406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,1024,0.011763555308183035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,768,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,768,0.011018666956159802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,512,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,256,0.002461333241727617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,512,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,512,0.010737777584128909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,3584,0.008290666672918532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,128,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,256,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,256,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,10240,0.03496266735924615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,256,0.009639999932712978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,128,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,32,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,128,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,65536,0.023772444989946153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,128,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,64,0.004960888789759742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,51200,0.017872000734011333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,32,0.005122666557629903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,1536,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,65536,0.07644710938135783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,65536,0.17716266049279106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,16384,0.01018666641579734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,51200,0.04810400141610039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,512,768,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,51200,0.13927643828921848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,12288,0.012069332930776807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,16384,0.017961778574519686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,16384,0.02237866653336419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,16384,0.05048888921737671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,10240,0.010610666539933948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,12288,0.01807733376820882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,512,64,0.0021199999998013177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,12288,0.039408889081743025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,8192,0.009302222066455418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,10240,0.013762666947311826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,10240,0.015867556134859722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,10240,0.03457066747877333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,7168,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,8192,0.011760888828171624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,65536,0.05884622202979194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,8192,0.029158221350775823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,7168,0.011040888726711273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,6144,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,7168,0.012635555532243518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,7168,0.026751110951105755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,5120,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,6144,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,6144,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,4096,0.006656000183688269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,6144,0.0239955551094479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,5120,0.00958933350112703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,12288,0.015107555521859063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,3584,0.006219555520349079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,5120,0.021752888957659405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,4096,0.009304888546466827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,4096,0.018920888503392536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,3072,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,3584,0.00831822223133511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,3584,0.008949333594905006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,3584,0.01755644381046295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,2560,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,8192,0.013668444421556262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,3072,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,3072,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,2048,0.004541333350870344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,2560,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,3072,0.01648888819747501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,51200,0.05926399760776096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,2560,0.00742222203148736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,1536,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,2560,0.014920888675583733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,2048,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,1024,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,2048,0.006784889019197888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,1536,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,2048,0.01387644476360745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,1536,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,768,0.002847111059559716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,1536,0.012424888710180918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,1024,0.00638755535085996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,5120,0.01036977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,1024,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,4096,0.008828444613350762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,768,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,768,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,256,0.0022088889446523455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,768,0.01070844464831882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,512,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,512,0.005392000079154968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,128,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,512,0.010682666467295753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,256,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,256,0.00508977762526936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,32,0.002161777681774563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,256,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,128,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,128,0.004874666531880696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,128,0.007698666718271043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,64,0.004976888911591636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,51200,0.013571555415789286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,256,32,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,65536,0.05804977814356486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,65536,0.0761937763955858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,16384,0.00923822240697013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,65536,0.17563111252254912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,51200,0.04483822319242689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,51200,0.05844089057710436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,12288,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,51200,0.13849688900841609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,256,512,0.002475555572244856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,16384,0.02199999988079071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,16384,0.018223999275101554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,1024,0.011415111521879831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,10240,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,12288,0.01511111193233066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,16384,0.04925955666436089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,12288,0.017886221408843994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,8192,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,10240,0.013804444836245643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,12288,0.03942311141226027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,10240,0.015847999188635085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,7168,0.0068897778789202375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,10240,0.034232000509897866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,8192,0.013482666677898832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,65536,0.01816177699300978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,6144,0.0063031112982167145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,8192,0.029161777761247423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,7168,0.011023110813564725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,7168,0.012418666647540199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,7168,0.02676888969209459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,6144,0.01037333326207267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,6144,0.011211555865075854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,6144,0.023824888798925612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,5120,0.009635555247465769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,5120,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,5120,0.021336888273557026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,3584,0.006092444476154115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,4096,0.009091555244392818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,4096,0.018889778190188937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,3584,0.00867911097076204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,3072,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,3584,0.01720977822939555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,8192,0.011742221812407175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,3072,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,2560,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,5120,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,3072,0.016193777322769165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,2048,0.004233777936961916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,2560,0.00757155567407608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,4096,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,2560,0.014819555812411837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,1536,0.003850666599141227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,2048,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,2048,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,4096,0.008708444734414419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,2048,0.01354311075475481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,1024,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,1536,0.0068986668354935115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,1536,0.006311999840868845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,768,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,1536,0.012406222522258759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,3584,0.008329778081840938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,1024,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,512,0.0024622221373849446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,1024,0.011736888852384357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,768,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,768,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,256,0.002151111140847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,768,0.010402667025725046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,512,0.005410666680998272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,512,0.010204444328943888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,128,0.0019964445382356644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,256,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,256,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,2560,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,128,0.005686222265164058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,128,0.004945777770545748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,128,32,0.001806222316291597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,65536,0.015204444527626038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,64,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,51200,0.01275466630856196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,128,32,0.004951999833186467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,16384,0.006994666324721442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,65536,0.07479733228683472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,1024,0.0063164445261160536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,12288,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,10240,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,51200,0.05859555800755819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,8192,0.005748444547255834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,12288,0.01759466694460975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,512,0.00612088872326745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,7168,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,10240,0.015643555257055495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,6144,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,5120,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,8192,0.013444444371594323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,4096,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,7168,0.012429333395428128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,3584,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,6144,0.011359111302428775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,3072,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,5120,0.010409778190983666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,2560,0.004967111266321606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,128,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,4096,0.009111111362775167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,2048,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,1536,0.003536000019974179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,3584,0.008793777889675563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,1024,0.0031724443866146933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,3072,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,768,0.002647111192345619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,2560,0.007329777710967594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,512,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,2048,0.00665511108107037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,256,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,1536,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,128,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,1024,0.0058755555914507965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,768,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,64,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,16384,0.021773333350817364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,512,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,64,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,256,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,3072,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,65536,0.026629333694775898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,128,0.004949333353175057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,16384,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,64,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,12288,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,65536,0.0748222205373976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,256,0.009368889033794403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,51200,0.05802400244606865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,8192,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,16384,0.021705778108702764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,7168,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,12288,0.01761155492729611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,6144,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,10240,0.015466666883892484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,5120,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,7168,0.012416000167528788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,4096,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,6144,0.011366222467687396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,3584,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,5120,0.010277333358923594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,3072,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,4096,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,2560,0.005035555611054103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,3584,0.008622222476535374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,51200,0.021664000219768945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,3072,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,1536,0.0035075553589397003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,64,32,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,2560,0.007355555891990662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,1024,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,768,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,512,0.0024604443460702896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,10240,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,2048,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,256,0.00222044438123703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,128,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,1536,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,64,0.0017724444882737265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,1024,0.005708444449636672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,768,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,32,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,512,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,256,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,8192,0.013436444103717804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,128,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,64,0.004859555512666702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,512,32,32,0.004865777575307422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,16384,0.40401776631673175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,12288,0.3049857881334093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,512,32,2048,0.004600000050332812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,16384,0.41341421339246964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,16384,0.2157804436153836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,10240,0.276239103741116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,12288,0.16798310809665254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,12288,0.3155946731567383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,51200,1.320757336086697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,10240,0.2567591137356228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,10240,0.1373644404941135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,8192,0.22813333405388725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,51200,1.3277787102593315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,7168,0.22126577960120308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,6144,0.17071110672420928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,8192,0.21229155858357748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,8192,0.10819111267725627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,7168,0.11115911271837021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,5120,0.14458221859402126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,7168,0.18136178122626412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,51200,0.6956604321797689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,4096,0.1272346708509657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,6144,0.15681333012051052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,5120,0.07039555576112536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,5120,0.13269955582088894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,6144,0.11970578299628364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,4096,0.05928711096445719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,3072,0.09203466441896226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,3584,0.12230222755008274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,3584,0.0616133345497979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,2560,0.07703910933600532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,4096,0.10855733023749457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,3584,0.09660622146394517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,3072,0.05426311161783007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,2048,0.062368002202775746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,1536,0.052770667605929904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,1024,0.03558933403756883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,3072,0.08277955320146349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,2560,0.052016887399885386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,2560,0.07153511047363281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,768,0.035190221336152815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,2048,0.05996355745527479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,512,0.020483555065260995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,1024,0.022264000442292955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,1024,0.036678221490648054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,768,0.027349332968393963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,768,0.029900444878472224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,256,0.017754667335086398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,512,0.01827911039193471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,128,0.01020088874631458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,512,0.02534399926662445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,256,0.017496888836224873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,64,0.009628444082207149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,256,0.0221404449807273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,128,0.01664799948533376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,65536,32,0.012073777616024017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,128,0.02163644466135237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,64,0.021272889441914026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,32,0.02132266594303979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,2048,0.03558399942186143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,65536,1536,0.028190222051408555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,65536,1536,0.04813955558670891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,16384,0.34160444471571183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,51200,1.0409439934624565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,65536,1.367335107591417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,16384,0.3377431233723958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,12288,0.26227111286587185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,51200,1.0518240398830838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,16384,0.1706613302230835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,10240,0.21002754900190565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,65536,0.6862097846137153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,12288,0.12967999776204428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,51200,0.5405084292093912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,12288,0.24268176820543078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,10240,0.2012711101108127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,10240,0.108897778722975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,7168,0.16277421845330134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,8192,0.1018222239282396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,6144,0.13079910808139378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,7168,0.14342489507463244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,8192,0.1731884479522705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,65536,1.3626293606228297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,8192,0.16983289188808867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,5120,0.10885333352618748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,6144,0.14316622416178384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,7168,0.08814755413267349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,4096,0.09011644124984741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,6144,0.0677715539932251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,3584,0.08439822329415216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,5120,0.05811022387610542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,3072,0.07586489121119182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,5120,0.10503200027677749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,2560,0.06392711400985718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,4096,0.08653244707319473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,4096,0.04792622062895033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,3072,0.04336799846755134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,3584,0.0762959983613756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,1536,0.03911555475658841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,3072,0.06702044275071886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,2560,0.03335377905103896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,2048,0.029045333464940388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,1024,0.028445333242416382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,3584,0.041895111401875816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,1536,0.03913155529234145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,768,0.02163022259871165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,2560,0.057894223266177706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,1024,0.030228445927302044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,512,0.01550222271018558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,1536,0.023144889209005568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,768,0.017027555240525138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,256,0.011027555498811932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,768,0.024488889508777197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,1024,0.018744889232847426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,128,0.009119999905427298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,512,0.015829333000712924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,512,0.020407110452651978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,64,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,256,0.014366222752465142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,32,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,51200,128,0.014317333698272705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,128,0.01757066614098019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,32,0.01758044461409251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,64,0.017575111654069688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,51200,2048,0.04709688822428385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,65536,0.3949315547943115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,2048,0.04857244425349765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,51200,0.31054843796624076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,16384,0.105185780260298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,65536,0.44566043217976886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,65536,0.23736800087822807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,12288,0.08011111285951403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,51200,0.1887164380815294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,51200,0.3414711157480876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,16384,0.061958220269944936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,16384,0.11222399605645074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,12288,0.04826755656136406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,10240,0.041343109475241765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,10240,0.0716257757610745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,7168,0.05242310961087545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,51200,256,0.01853155593077342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,8192,0.05910578038957384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,6144,0.04579910967085096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,7168,0.030588444736268785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,7168,0.05217422379387749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,6144,0.030717333157857258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,6144,0.04575733343760172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,5120,0.03931644558906555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,5120,0.02389066749148899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,4096,0.03343910972277323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,4096,0.020409777760505676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,10240,0.07264710797203912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,3584,0.02942399846182929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,4096,0.032607999112870954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,8192,0.06134400102827284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,3072,0.025607110725508794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,3584,0.018789332773950364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,2560,0.021873777111371357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,8192,0.03988800115055508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,3072,0.01718844473361969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,3584,0.029221332735485498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,2048,0.017602667212486267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,3072,0.025639111797014873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,2560,0.01567555632856157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,1536,0.01388177772363027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,5120,0.03956088754865859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,2048,0.014067555467287699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,2560,0.02295733326011234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,1024,0.010704889065689512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,2048,0.01928444372283088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,768,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,1536,0.012434666355450949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,512,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,1536,0.016400888562202454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,1024,0.010325333310498131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,256,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,768,0.009383999639087254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,1024,0.01346577786737018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,128,0.0042444442709287005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,768,0.011525332927703857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,512,0.009679111341635386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,256,0.008299555215570662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,64,0.004226666771703296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,256,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,128,0.008404444489214156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,128,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,12288,0.08445244365268284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,64,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,16384,32,0.008989333278603023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,65536,0.29860443539089626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,51200,0.23772533734639487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,65536,0.3466782304975722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,16384,0.07803644736607869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,12288,0.060552000999450684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,51200,0.26953689257303876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,16384,0.05244355400403341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,16384,0.08732266558541192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,51200,0.16360355748070612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,10240,0.052087999052471586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,8192,0.042832887834972806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,12288,0.06681155496173434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,16384,32,0.004504888835880491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,10240,0.057145780987209745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,7168,0.03757244348526001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,8192,0.027181333965725366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,8192,0.04717866579691569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,6144,0.032680001523759626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,7168,0.024494222468800012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,7168,0.04211289021703932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,5120,0.028190222051408555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,6144,0.021938666701316833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,5120,0.019646222392718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,6144,0.036738667223188616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,65536,0.18183910846710205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,5120,0.03192266821861267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,3584,0.021609778205553692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,4096,0.017539555827776592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,3072,0.018735110759735107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,4096,0.02624977793958452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,3584,0.023754666248957317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,3584,0.01594311164485084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,12288,0.038839111725489296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,2560,0.01609866652223799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,2048,0.013789333403110504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,10240,0.033949332104788885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,3072,0.021198223034540813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,1536,0.011233777635627322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,2560,0.01898755629857381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,2048,0.012107555237081317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,1024,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,2048,0.016487111647923786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,1536,0.010618666807810465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,768,0.007925333248244392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,1024,0.009103111094898647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,1024,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,512,0.0063582224150498705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,768,0.008554666406578487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,768,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,4096,0.024822221861945257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,256,0.004528000122971005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,512,0.008035555481910706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,128,0.00388711111413108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,256,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,512,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,64,0.0034977777136696708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,256,0.008083555433485243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,16384,512,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,128,0.007624000310897827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,12288,32,0.0036026665733920205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,128,0.007709333466158972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,3072,0.0148062225845125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,12288,2560,0.013595555391576556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,64,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,32,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,65536,0.3030924532148573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,51200,0.2207191122902764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,12288,1536,0.013952000273598565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,65536,0.168449772728814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,65536,0.2908257908291287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,51200,0.23148798942565918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,12288,0.061899556054009326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,51200,0.13037333223554823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,16384,0.0464915566974216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,10240,0.0506640010409885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,16384,0.07657155725691053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,8192,0.040692445304658674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,12288,0.03722755445374383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,12288,0.05875466929541694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,7168,0.03596622082922194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,10240,0.05010044574737549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,8192,0.02609155575434367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,8192,0.041645334826575384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,7168,0.023711111810472276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,7168,0.03681688838534885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,5120,0.02743822170628442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,6144,0.02139288849300808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,6144,0.0322728885544671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,4096,0.023060444328520033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,5120,0.018872888551818002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,3584,0.02013777858681149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,5120,0.027549333042568628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,4096,0.016723554995324876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,3072,0.01792711185084449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,4096,0.023006222314304773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,3584,0.015407999356587728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,2560,0.015542222393883599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,3584,0.02136444383197361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,3072,0.014383110735151501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,2048,0.012834666503800286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,2560,0.012753778033786349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,2560,0.01718133356836107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,1536,0.011150221857759686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,2048,0.011582222249772815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,2048,0.014929778046078153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,6144,0.03211288981967502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,1024,0.009293333523803288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,1536,0.010092444717884064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,1536,0.012770666844315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,768,0.007917332980367873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,1024,0.008740444150235917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,512,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,1024,0.01034577770365609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,768,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,256,0.004408888932731417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,512,0.00776800016562144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,768,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,128,0.003866666720973121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,256,0.007370666497283512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,16384,0.0775795578956604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,64,0.0035084444615576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,256,0.007625777688291337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,10240,32,0.003866666720973121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,128,0.007481777833567725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,128,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,3072,0.019419555862744648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,64,0.007348444726732042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,65536,0.24221332867940268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,32,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,16384,0.06390400065316094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,10240,10240,0.0314648879898919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,65536,0.24494043986002603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,65536,0.1233288844426473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,12288,0.04979022343953451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,16384,0.03581510980923971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,51200,0.19272889031304252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,10240,0.042131556404961475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,16384,0.06487555636299981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,12288,0.02843022346496582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,8192,0.03490044342146979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,10240,0.024567999773555334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,10240,0.04295911060439216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,8192,0.020949333906173706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,10240,512,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,8192,0.03564533260133531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,6144,0.02658933401107788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,7168,0.021689777572949726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,5120,0.02293599976433648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,7168,0.03144888745413886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,6144,0.027797334724002417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,6144,0.017320889565679762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,5120,0.015790222419632804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,4096,0.019151111443837483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,3584,0.01657511128319634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,51200,0.18967200650109184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,5120,0.023982221881548565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,3072,0.014648000399271647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,4096,0.014075555735164218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,4096,0.020388444264729817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,51200,0.10030311346054077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,3584,0.013086222112178802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,2560,0.012538666526476542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,3584,0.018568888306617737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,2048,0.010671111444632212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,3072,0.011810666984981961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,3072,0.017096888687875535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,1536,0.009004444711738164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,2560,0.010979555547237396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,2560,0.015246222416559855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,1024,0.0069955554273393415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,7168,0.030574222405751545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,1536,0.008959111240175035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,2048,0.013232889274756113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,1536,0.011117333339320289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,1024,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,512,0.004513777792453766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,1024,0.009006222089131674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,768,0.007679999702506595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,256,0.0035537779331207275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,768,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,512,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,128,0.0031866667171319327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,512,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,256,0.006719999843173557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,64,0.003264888914095031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,256,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,128,0.006912888752089606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,128,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,32,0.0032986667421129015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,64,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,32,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,65536,0.23219733768039277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,8192,2048,0.00963644435008367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,65536,0.20955555968814424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,51200,0.1821919944551256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,65536,0.12045066886478001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,8192,768,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,12288,0.047510223256217114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,51200,0.16506044069925943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,51200,0.09493511252933079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,16384,0.03619911273320516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,10240,0.04068266683154636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,12288,0.02721066607369317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,8192,0.034091555409961276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,12288,0.04278666774431864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,10240,0.023806222611003455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,10240,0.036944889360004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,7168,0.029475556479560003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,8192,0.020396444532606337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,8192,0.030505776405334473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,7168,0.01863555610179901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,5120,0.021869333253966436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,7168,0.027169777287377253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,8192,12288,0.050266666544808276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,6144,0.01737511157989502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,6144,0.024152888192070857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,4096,0.018223111828168232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,5120,0.015494222442309061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,3584,0.015980444020695157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,5120,0.020672889219390023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,4096,0.013790222505728403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,3072,0.01424888935354021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,16384,0.0636791123284234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,4096,0.01805688937505086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,3584,0.012785777449607849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,2560,0.012337778177526263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,2048,0.010619555910428366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,3584,0.016490666402710807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,3072,0.011485333244005838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,1536,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,2560,0.010703999963071613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,3072,0.015174221661355762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,16384,0.055218666791915894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,2048,0.009479111267460717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,2560,0.013538666897349887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,1024,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,768,0.00563822231358952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,2048,0.011940444509188334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,1536,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,512,0.004552000098758274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,1024,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,1536,0.009863999982674917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,256,0.003931555483076307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,768,0.0074728892909155945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,128,0.003548444559176763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,512,0.0069822221994400024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,256,0.006683555742104848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,768,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,512,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,64,0.0031786666562159858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,32,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,7168,128,0.006718222051858902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,128,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,64,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,32,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,65536,0.19554666678110758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,51200,0.1503573391172621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,65536,0.19817154937320283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,65536,0.11663111050923665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,16384,0.05538488758934868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,51200,0.09237866931491429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,51200,0.15506933795081243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,12288,0.04297511114014519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,16384,0.033653332127465144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,16384,0.05344355437490675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,10240,0.035263111193974815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,12288,0.026576000783178542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,8192,0.029197331931855943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,1024,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,10240,0.02334844403796726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,10240,0.03560444381501939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,7168,6144,0.025729777084456548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,8192,0.020059555768966675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,7168,256,0.006607999818192587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,6144,0.021576889687114294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,8192,0.029459555943806965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,5120,0.018775999546051025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,7168,0.018360889620251127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,7168,0.02658133374320136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,6144,0.018043556147151522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,4096,0.015139554937680563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,6144,0.023559111687872145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,5120,0.015207111835479736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,3584,0.013231111069520315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,5120,0.020440888073709276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,4096,0.013442666994200813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,3072,0.012177777787049612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,4096,0.017645332548353408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,3584,0.01293688929743237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,2560,0.010292444792058732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,3072,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,3584,0.016148444679048326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,2048,0.009045333498054082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,3072,0.014833778142929077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,2560,0.010676444404655032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,1536,0.0075777777367168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,2560,0.013269333375824822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,2048,0.009335999687512716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,1024,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,12288,0.04157155421045091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,2048,0.011345778074529437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,768,0.005097777893145879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,512,0.0038408889538711975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,1536,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,7168,0.02537244392765893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,256,0.0031946665710873077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,768,0.0074648890230390764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,1024,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,128,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,512,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,768,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,64,0.0026559999419583213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,256,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,6144,32,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,512,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,128,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,256,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,128,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,64,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,65536,0.18789154953426787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,6144,32,0.006566222343179915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,51200,0.15093689494662815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,16384,0.050091554721196495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,65536,0.11337510744730632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,65536,0.16266666518317327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,12288,0.03909066650602553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,51200,0.08998044331868489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,16384,0.033273776372273765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,51200,0.1267484426498413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,10240,0.033596442805396184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,12288,0.025736888249715168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,16384,0.04359111189842224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,8192,0.027288888891537983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,1536,0.008400888906584846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,12288,0.03371289041307237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,7168,0.0239955551094479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,10240,0.022575111852751836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,6144,0.020915556285116408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,8192,0.0200284438000785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,10240,0.029183109601338703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,7168,0.01862844493654039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,8192,0.02446844511561924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,4096,0.01462399959564209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,5120,0.014460444450378418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,6144,0.019996444384257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,5120,0.017422222428851657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,3584,0.013069333301650153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,4096,0.012522666818565793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,4096,0.01513777838812934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,3072,0.012065778175989786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,3584,0.011984888878133563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,2560,0.010273777776294285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,3584,0.01391999920209249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,3072,0.010504888991514841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,2048,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,3072,0.012814222110642327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,2560,0.01038755559258991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,2560,0.011071110765139261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,1536,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,2048,0.008952889177534316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,2048,0.009681777821646797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,1024,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,1536,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,6144,1024,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,1536,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,768,0.004807999978462855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,1024,0.007692444655630324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,512,0.0038693332009845306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,6144,0.016354666815863717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,1024,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,256,0.00313688897424274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,768,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,768,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,512,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,128,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,512,0.006313777632183499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,256,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,64,0.002528888897763358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,128,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,128,0.005971555494599872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,64,0.005966222120655908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,65536,0.1312862237294515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,32,0.006079999936951532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,51200,0.10563911332024468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,65536,0.06865600082609388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,65536,0.15396800306108263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,5120,0.01776266594727834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,16384,0.042543109920289784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,5120,7168,0.02237244447072347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,51200,0.12002400557200114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,16384,0.021465778350830078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,5120,256,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,10240,0.024705777565638225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,12288,0.017677333619859483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,12288,0.03222399950027466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,8192,0.02057422200838725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,10240,0.016061334146393668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,10240,0.028015110227796767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,7168,0.017860444055663217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,8192,0.013953778478834363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,8192,0.02346844474474589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,6144,0.015841777125994366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,7168,0.014375110467274984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,5120,0.013801777528391944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,6144,0.013080000049538083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,7168,0.021538666552967493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,4096,0.011384000380833944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,6144,0.0192248887485928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,5120,0.010407999985747868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,3584,0.010661332971519895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,5120,0.016859556237856548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,51200,0.05496266815397474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,4096,0.009087111386987898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,3072,0.009573332965373993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,4096,0.014550222290886773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,16384,0.04174400038189358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,3584,0.008771555291281806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,3584,0.013440888788965015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,2048,0.007361777954631382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,5120,32,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,3072,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,2560,0.008375110725561777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,1536,0.006015111174848344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,2560,0.010496888723638324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,2048,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,1024,0.004951111144489712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,1536,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,2048,0.009380444056457942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,768,0.004224888980388641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,1536,0.008268444074524773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,1024,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,512,0.0035386664999855887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,1024,0.007290666302045186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,768,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,256,0.0031884445084465873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,512,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,128,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,512,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,256,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,64,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,256,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,128,0.005994666781690385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,12288,0.029398222764333088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,32,0.0025475554996066624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,128,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,4096,2560,0.008399110701349046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,64,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,4096,3072,0.008307555483447181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,32,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,51200,0.10109688838322957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,65536,0.06703911225001018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,16384,0.04090133309364319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,65536,0.11775288316938613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,12288,0.028112000889248315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,51200,0.05387377738952637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,16384,0.020909332566791113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,51200,0.09240000115500556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,10240,0.02370311154259576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,16384,0.03184088733461168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,8192,0.01941866676012675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,12288,0.025528889563348558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,10240,0.01568888955646091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,10240,0.022543110781245764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,7168,0.01720177796151903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,4096,768,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,8192,0.013753777576817406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,6144,0.015455111861228943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,8192,0.01885066595342424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,5120,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,7168,0.012786666552225748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,6144,0.011232888533009423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,4096,0.011593778100278644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,6144,0.015847111741701763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,5120,0.010055111514197456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,3584,0.010310221877362993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,5120,0.01385600037044949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,65536,0.12583111392127141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,4096,0.008981333010726506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,3072,0.009391110804345872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,3584,0.008795555267069075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,2560,0.0081386665503184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,3584,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,3072,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,2048,0.007819555699825287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,3072,0.010577778021494547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,2560,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,12288,0.017468444175190397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,1536,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,2048,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,1024,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,2048,0.00830844458606508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,1536,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,768,0.0042408886882993905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,1536,0.007368000017272101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,1024,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,512,0.0035297779573334586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,1024,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,7168,0.01754488878779941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,256,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,512,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,512,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,128,0.0026177778426143858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,256,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,4096,0.012058667010731168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,256,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,64,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3584,32,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,128,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,128,0.00553688903649648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,65536,0.10997422536214192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,64,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,2560,0.009240888886981541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,32,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,65536,0.06622222397062513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,16384,0.032568001084857516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,65536,0.11706754896375869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,12288,0.025408888856569927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,16384,0.020583111378881667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,51200,0.09092799822489421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,16384,0.03175110949410333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3584,768,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,12288,0.017109332813156974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3584,768,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,8192,0.017816889617178176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,12288,0.02535022298494975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,10240,0.015440889530711703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,10240,0.02218222286966112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,8192,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,6144,0.013849777479966482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,7168,0.012314666476514606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,7168,0.017356443736288283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,5120,0.012065778175989786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,6144,0.011063999599880643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,6144,0.015815999772813585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,4096,0.01016177816523446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,5120,0.009764444496896531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,51200,0.08696799808078343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,3584,0.008951110972298516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,5120,0.01386577801571952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,4096,0.009004444711738164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,4096,0.011687111523416309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,3072,0.008150222400824228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,51200,0.05265511075655619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,3584,0.008383110993438298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,2560,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,10240,0.02214844524860382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,3072,0.00831911133395301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,2048,0.006215111249023014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,2560,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,3072,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,1536,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,2560,0.008835555778609382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,1024,0.004045333299371931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,2048,0.008319999608728621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,1536,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,768,0.003539555602603488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,1536,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,8192,0.018989332848125033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,1024,0.006681777950790193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,512,0.0032337777730491427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,768,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,768,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,256,0.002551111082235972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,512,0.006058666855096817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,512,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,128,0.0024968888610601425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,256,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,128,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,3584,0.010788444843557147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,128,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,32,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,2048,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,7168,0.015442666080262927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,64,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,65536,0.08942044443554348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,32,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,3072,1024,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,65536,0.06511466370688544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,16384,0.03014755580160353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,51200,0.051742222574022084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,51200,0.08695111009809707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,12288,0.02500177754296197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,16384,0.020272889071040683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,16384,0.030846221579445735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,10240,0.02032444377740224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,3072,256,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,12288,0.016861332787407767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,8192,0.01685244507259793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,12288,0.024878222081396315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,10240,0.015289778510729471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,7168,0.015068444940778943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,3072,64,0.002463111033042272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,10240,0.021924444370799597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,6144,0.013071111506885953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,8192,0.018562666243977018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,7168,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,5120,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,7168,0.017142222987280954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,4096,0.010076444182131026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,6144,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,6144,0.015370666980743408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,5120,0.009700444837411245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,3584,0.009534222384293875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,5120,0.013434666726324292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,4096,0.008987555901209513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,4096,0.011167111496130625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,3584,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,65536,0.11113688680860732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,2560,0.007345777418878343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,3584,0.010393777655230628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,3072,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,2048,0.006405333264006509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,2560,0.008023999631404877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,1536,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,2560,0.008949333594905006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,2048,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,1024,0.004408888932731417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,2048,0.008037333687146505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,1536,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,768,0.003842666745185852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,1536,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,1024,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,8192,0.012730666332774691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,512,0.0033004445334275565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,1024,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,256,0.0027715555495686005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,512,0.006115555349323485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,768,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,512,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,128,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,256,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,256,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,51200,0.07249689102172852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,64,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,3072,0.008468444148699442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2560,32,0.0022035555707083806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,128,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,65536,0.07206844621234469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,51200,0.05694222450256348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,32,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,3072,0.009712888962692684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,65536,0.0634853310055203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,16384,0.023327999644809302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,65536,0.10622311300701565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,51200,0.05029511120584276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,12288,0.018529777725537617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,16384,0.019912888606389362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,16384,0.03021777669588725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,12288,0.016301333904266357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,12288,0.024115555816226538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,768,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,10240,0.01443199978934394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,10240,0.0210817770825492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,8192,0.012099555797047086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,7168,0.015203555425008139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,8192,0.018056000272432964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2560,128,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,7168,0.011537777880827585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,7168,0.016340444485346477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2560,64,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,6144,0.01036800030204985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,6144,0.01443288889196184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,5120,0.00962488849957784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,4096,0.010032888915803697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,51200,0.08378399742974176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,5120,0.012410666379663678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,10240,0.02039377722475264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,3584,0.007810666329330868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,4096,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,3072,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,8192,0.01724088854259915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,4096,0.010859555668301053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,2560,0.005991111199061076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,3584,0.010173333187898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,3072,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,2048,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,3072,0.009367999931176504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,2560,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,1536,0.004543999830881755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,6144,0.013767999907334646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,2048,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,1024,0.0038720000949170855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,1536,0.006948444578382704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,1536,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,5120,0.012058667010731168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,768,0.0033200000309281876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,1024,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,1024,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,768,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,768,0.006129777679840724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,256,0.00249422217408816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,512,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,512,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,128,0.002201777779393726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,3584,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,256,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,64,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,256,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,32,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,2048,128,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,128,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,65536,0.05806666612625122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,2560,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,64,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,51200,0.04742222362094455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,32,0.005722666780153911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,65536,0.06158044603135851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,65536,0.06641688611772326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,51200,0.0487937761677636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,12288,0.014783999986118741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,16384,0.0196115556690428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,51200,0.05247111121813456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,10240,0.015832000308566622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,2048,512,0.002846222163902389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,12288,0.015855110353893705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,12288,0.017041777571042378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,10240,0.01349688900841607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,10240,0.014941333068741692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,8192,0.011696889168686338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,8192,0.012800888882742988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,6144,0.010839111275143094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,7168,0.011035555766688453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,7168,0.01181333346499337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,5120,0.009687999884287516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,6144,0.010061333576838175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,6144,0.010684444672531553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,4096,0.008683555656009251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,5120,0.009658666948477427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,16384,0.017895999881956313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,2048,2048,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,3584,0.007714666426181793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,4096,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,4096,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,3584,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,16384,0.020336889558368258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,3584,0.008047111332416534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,2560,0.0058115555180443665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,3072,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,8192,0.01531644496652815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,2048,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,3072,0.007663110891977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,2560,0.007589333587222629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,7168,0.013821333646774292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,1536,0.0041884444653987885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,2048,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,1024,0.0034968890249729156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,2560,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,768,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,2048,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,1536,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,512,0.0028524444335036804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,1536,0.006311111152172089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,768,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,1024,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,256,0.0025271111064487034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,768,0.005678222411208683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,512,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,512,0.0053084443012873335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,256,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,256,0.00518488883972168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,5120,0.009373333719041612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,64,0.002104000084929996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,32,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,128,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,65536,0.041095998552110456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,128,0.005033777819739448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,3072,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,51200,0.032912890116373696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,32,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,65536,0.05939110782411364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,16384,0.014520888527234396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,65536,0.06392888890372382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,51200,0.04793955551253425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,51200,0.050331556134753756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,16384,0.01805688937505086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,16384,0.01994755533006456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,10240,0.012077333198653327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,12288,0.014800000521871777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,12288,0.015846222639083862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1536,128,0.002161777681774563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,8192,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,10240,0.013536888692114087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,7168,0.011385777758227454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,8192,0.011716444459226398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,8192,0.012080888781282636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,6144,0.010076444182131026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,7168,0.011039999624093374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,7168,0.011423110961914062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,5120,0.008721777962313758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1536,64,0.005062222066852781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,6144,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,4096,0.0070471109615431885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,6144,0.010403555300500657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,3584,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,4096,0.008635555704434713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,12288,0.012238221863905588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,4096,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,3072,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,3584,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,2560,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1536,1024,0.0064080001579390625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,3072,0.007914666500356462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,2048,0.004528000122971005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,3072,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,2560,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,1536,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,2560,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,10240,0.014263111684057446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,1024,0.0032097777972618737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,1536,0.006855999843941794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,1536,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,768,0.0028773333049482773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,1024,0.0063528890411059065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,1024,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,512,0.002528888897763358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,768,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,768,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,5120,0.009391110804345872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,256,0.0022115556316243275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,5120,0.009332444104883406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,512,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,128,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,512,0.0052826665341854095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,256,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,64,0.0020702222569121257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,1024,32,0.0021564445147911706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,256,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,3584,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,65536,0.03303555647532145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,128,0.005008888741334279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,64,0.005043555465009478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,51200,0.025703110628657874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,32,0.004958222309748332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,65536,0.0588702228334215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,16384,0.013032889200581444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,65536,0.06222489145067003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,12288,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,16384,0.017882666654056974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,51200,0.048877331945631236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,16384,0.019239111079110038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,10240,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,12288,0.014695111248228284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,12288,0.015511110424995422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,8192,0.00906399968597624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,10240,0.013450666434235044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,7168,0.010025777750545079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,8192,0.011557333171367645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,10240,0.013823111024167804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,6144,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,8192,0.012074666718641916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,7168,0.010936000280910067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,5120,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,128,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,7168,0.011207111179828644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,4096,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,6144,0.010339555641015371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,5120,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,5120,0.009356444080670675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,4096,0.008627555436558193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,3584,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,4096,0.00834222220712238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,1024,2048,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,3072,0.0058293334311909145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,1024,2048,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,3584,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,2560,0.005193777796294955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,3584,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,3072,0.00793333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,2048,0.004523555437723796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,2560,0.007595555649863348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,3072,0.0075902218619982404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,1536,0.0038497779104444715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,2560,0.007012444237867991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,2048,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,1024,0.0032097777972618737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,1536,0.00684799998998642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,2048,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,768,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,1536,0.006301333506902059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,512,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,768,0.006215111249023014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,1024,0.005701333284378052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,256,0.0022275555464956495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,768,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,6144,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,512,0.005883555445406172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,128,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,512,0.005425777700212266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,256,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,64,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,768,32,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,128,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,256,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,65536,0.024958221448792353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,128,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,51200,0.02176977859603034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,51200,0.046894222497940063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,64,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,768,32,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,16384,0.012888000243239932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,65536,0.05918755796220568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,12288,0.012709333664841123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,65536,0.06090933084487915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,51200,0.0451564457681444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,16384,0.017281777328915067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,10240,0.011295999917719098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,51200,0.04850222335921394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,12288,0.014489778214030795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,8192,0.009959110783206092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,12288,0.01514488955338796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,10240,0.013424889081054263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,7168,0.008845333423879411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,8192,0.011385777758227454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,768,1024,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,6144,0.008277333445019191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,10240,0.013424889081054263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,8192,0.011991110940774282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,7168,0.010830221904648675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,6144,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,7168,0.011239999698268043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,4096,0.0069066666894488865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,6144,0.01036977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,3584,0.00630044440428416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,5120,0.009199110998047723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,4096,0.008566222257084316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,3072,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,4096,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,2560,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,3072,0.007620444728268518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,3072,0.007595555649863348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,2048,0.004550222307443619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,2560,0.007375111182530721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,2560,0.00699733363257514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,1536,0.0038613333470291565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,2048,0.006994666324721442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,16384,0.018917333748605516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,2048,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,1536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,1536,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,768,0.0028168888141711554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,5120,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,1024,0.0063484443558586975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,1024,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,512,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,5120,0.009328888522254096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,768,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,256,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,512,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,768,0.005532444351249271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,128,0.0018613334331247543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,512,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,64,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,3584,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,128,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,256,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,65536,0.020121778051058452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,3584,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,128,0.00500177757607566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,51200,0.017140444782045152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,512,32,0.004944889081848992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,65536,0.05512977639834086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,16384,0.009898666706350114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,65536,0.05782310830222236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,51200,0.041793776883019336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,1024,0.003165333428316646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,12288,0.009663999908500248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,16384,0.01716711123784383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,51200,0.04639200038380093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,10240,0.008764444126023186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,16384,0.018234666850831773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,12288,0.014495111174053617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,8192,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,12288,0.015111999379263984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,10240,0.013080889152155982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,7168,0.008419555922349295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,10240,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,8192,0.011371555427710215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,6144,0.00793333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,8192,0.011932444241311816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,7168,0.010784000158309937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,5120,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,7168,0.011063111325105032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,512,256,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,4096,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,512,32,0.0018257778137922287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,6144,0.010010666317409938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,5120,0.009311999711725447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,5120,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,4096,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,3072,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,4096,0.008103999826643202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,3584,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,2560,0.004960888789759742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,3584,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,3072,0.007872888611422645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,2048,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,2560,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,1536,0.00370133337047365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,2560,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,2048,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,1024,0.0031591111587153543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,2048,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,1536,0.006700444552633498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,768,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,1024,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,1536,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,512,0.002491555487116178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,768,0.006228444476922353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,1024,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,256,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,768,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,128,0.0018959999498393799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,512,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,256,0.0058755555914507965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,256,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,128,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,3584,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,65536,0.015429332852363586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,128,0.004960888789759742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,64,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,51200,0.013205333716339536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,32,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,65536,0.05155644483036465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,16384,0.008985777695973715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,256,3072,0.00738844441043006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,51200,0.04173511266708374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,12288,0.011415111521879831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,16384,0.01717688971095615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,16384,0.018262222409248352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,10240,0.010254222485754225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,12288,0.014451555079884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,12288,0.014801777071423002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,8192,0.009015999734401703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,512,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,10240,0.013101333545313941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,7168,0.00831822223133511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,256,6144,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,8192,0.011345778074529437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,6144,0.007791111038790808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,256,32,0.0018275556051068837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,7168,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,5120,0.007071110937330458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,7168,0.010749333434634738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,6144,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,4096,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,6144,0.009909333454238044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,5120,0.009083555804358589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,3584,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,5120,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,65536,0.058487998114691846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,3072,0.0053137776752312975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,4096,0.008346666892369589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,4096,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,51200,0.0462337765428755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,3584,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,2560,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,3072,0.007720888488822513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,3584,0.007592000067234039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,2048,0.004202666795916027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,3072,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,2560,0.007376888559924231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,1536,0.0035964445107513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,2560,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,2048,0.006990222467316522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,1024,0.0030248889492617715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,1536,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,2048,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,768,0.002847111059559716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,10240,0.013456888496875763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,1024,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,1536,0.006140444427728653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,512,0.002528888897763358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,256,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,768,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,1024,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,8192,0.011736888852384357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,128,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,512,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,768,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,64,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,128,32,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,512,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,128,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,65536,0.019102222389645047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,256,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,128,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,51200,0.01681688924630483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,16384,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,64,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,128,32,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,10240,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,65536,0.05783377753363716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,51200,0.04552355408668518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,8192,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,7168,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,12288,0.01460355520248413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,6144,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,10240,0.013415111435784234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,5120,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,8192,0.011728888584507836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,4096,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,6144,0.00982488857375251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,3584,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,5120,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,3072,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,4096,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,2560,0.004618666652176115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,257,128,256,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,2048,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,3584,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,1536,0.0035573335157500375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,3072,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,1024,0.0028328889360030494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,12288,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,2560,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,768,0.002846222163902389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,512,0.0025048889219760895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,1536,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,256,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,16384,0.018233777748213876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,128,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,1024,0.00563822231358952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,768,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,64,0.0018115554832749898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,64,32,0.0018159999615616268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,512,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,65536,0.01991911066903008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,256,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,51200,0.017704889178276062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,128,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,64,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,12288,0.011680888632933298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,32,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,10240,0.010392888552612728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,65536,0.05644444624582926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,8192,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,51200,0.04521244433191088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,16384,0.018183999591403537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,7168,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,6144,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,10240,0.013316444224781461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,5120,0.006309333360857434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,8192,0.011448000040319232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,7168,0.011010666688283285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,6144,0.009723555710580613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,3584,0.005881777654091517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,2048,0.006304889089531369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,3072,0.0052844443255000645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,5120,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,2560,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,4096,0.008104888929261101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,2048,0.0041511112617121804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,3584,0.0075937774446275495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,1536,0.0035546666218174826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,3072,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,2560,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,64,7168,0.01073155552148819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,16384,0.014528888795110913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,2048,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,768,0.0027031112048361036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,512,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,1536,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,256,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,1024,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,768,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,128,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,12288,0.014848889576064216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,64,0.0017884444031450485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,512,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,32,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,256,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,4096,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,128,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,64,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,257,32,32,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,257,32,1024,0.0028728888266616394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,16384,0.3440728982289632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,51200,1.1063004599677193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,16384,0.14098400539822048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,51200,1.5272231631808812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,51200,0.4234328799777561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,12288,0.37590577867296004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,12288,0.25675998793707955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,16384,0.49621158175998265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,10240,0.2481093406677246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,12288,0.10811822281943427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,8192,0.22691554493374297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,10240,0.21338221761915419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,8192,0.18055200576782227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,10240,0.09108089076148139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,16384,0.3224284383985731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,8192,0.07359111309051514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,7168,0.19927733474307588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,7168,0.156713777118259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,6144,0.1598328881793552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,7168,0.06541066699557833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,12288,0.23921333418952095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,10240,0.20206844806671143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,8192,0.16318044397566053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,5120,0.1320755614174737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,6144,0.0572284460067749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,6144,0.1328808863957723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,4096,0.11606844266255696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,7168,0.1467519998550415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,5120,0.11126844088236491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,5120,0.049357334772745766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,3584,0.09332444270451863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,4096,0.09160977602005005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,4096,0.04224888814820183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,5120,0.10612089104122585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,6144,0.12477600574493408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,3072,0.08261777957280476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,3584,0.08126044273376465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,3584,0.03695111142264472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,4096,0.08435733450783624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,2560,0.0696444445186191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,3072,0.03364355696572198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,2048,0.059454222520192467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,2560,0.02960977620548672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,2560,0.06083466609319051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,2560,0.059431109163496226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,1536,0.04539822207556831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,2048,0.025822222232818604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,2048,0.05107466710938347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,2048,0.05126133229997423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,1024,0.03290311164326138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,1536,0.04074222180578444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,1536,0.020592888196309406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,1536,0.04327555497487386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,768,0.025126222107145522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,1024,0.01588711142539978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,1024,0.032072888480292425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,1024,0.034929778840806745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,512,0.018582221534517076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,768,0.014231999715169271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,3072,0.07099466853671603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,768,0.025830222500695124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,3072,0.06768622001012166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,256,0.01128533316983117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,768,0.030340443054835003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,512,0.013214222258991666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,512,0.021651556094487507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,128,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,512,0.02619022296534644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,256,0.012125333150227865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,64,0.00850044439236323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,51200,0.9415351019965278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,256,0.018938667244381376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,65536,32,0.012599999705950419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,256,0.020901333954599168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,128,0.011702222128709158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,128,0.01850755512714386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,128,0.016689777374267578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,64,0.018090666996108163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,65536,32,0.01791733337773217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,3584,0.07589244180255465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,51200,1.037009769015842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,51200,0.8494017918904623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,65536,1.1004915237426758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,65536,1.3710418277316625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,65536,0.4180782106187608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,16384,0.3307110998365614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,16384,0.2756800121731228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,12288,0.2532880041334364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,16384,0.1218035618464152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,12288,0.20395999484592012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,10240,0.17576355404324004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,12288,0.08459822336832683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,51200,0.338848008049859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,10240,0.17084177335103354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,10240,0.07247733407550387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,8192,0.14668177233801946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,16384,0.25859467188517254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,12288,0.1994933287302653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,7168,0.12813956207699245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,8192,0.05946489175160726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,10240,0.15866043832567003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,7168,0.12360889381832546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,8192,0.13919556140899658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,6144,0.11362577809227838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,7168,0.05295200149218241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,8192,0.13358488347795275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,5120,0.09409511089324951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,6144,0.045800887876086764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,7168,0.11774399545457627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,6144,0.10570310884051853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,4096,0.08726755777994792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,5120,0.0899626678890652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,5120,0.039292444785436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,3584,0.07409689161512586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,4096,0.0723759995566474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,4096,0.0340266658200158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,5120,0.0865031083424886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,3584,0.0637520021862454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,3072,0.05790222353405423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,4096,0.07233422332339816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,3584,0.06455022096633911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,3072,0.05598577525880602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,2560,0.04890666736496819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,3072,0.027434666951497395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,51200,0.7837546666463217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,2048,0.0414373344845242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,2560,0.04903466502825419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,2560,0.02442577812406752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,6144,0.10048088762495254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,1536,0.03164266546567281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,2560,0.050511999262703784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,2048,0.020655110478401184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,2048,0.045307556788126625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,1024,0.024296889702479046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,2048,0.044551998376846313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,1536,0.016918222109476726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,1536,0.03327733278274536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,65536,1.0423040390014648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,768,0.017998221847746108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,3584,0.030375109778510198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,1536,0.03754666778776381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,1024,0.013832888669437833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,512,0.013454222016864352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,768,0.012795555922720166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,1024,0.030563556485705908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,256,0.010695999695195092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,512,0.01752444439464145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,512,0.011725333001878528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,512,0.022669333550665114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,128,0.008948444492287105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,256,0.010400888820489248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,256,0.016057777735922072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,256,0.01919555498494042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,64,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,51200,128,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,51200,32,0.008656889200210571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,128,0.01516266663869222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,64,0.015159999330838522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,32,0.01517599986659156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,65536,0.39212621582878965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,1024,0.026521777113278706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,65536,0.3666222095489502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,3072,0.05765955315695869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,51200,0.306076447168986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,65536,0.14908356136745876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,128,0.015384000208642749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,16384,0.11565421687232123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,51200,0.2904088762071398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,51200,0.11575555801391602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,16384,0.04149333304829068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,16384,0.09362044599321152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,12288,0.0856364435619778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,51200,768,0.026738666825824316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,16384,0.09155199925104777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,12288,0.03267200125588311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,10240,0.06531377633412679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,12288,0.0710480014483134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,8192,0.05823644664552477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,10240,0.028370665179358587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,65536,0.3318213356865777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,51200,0.2682222260369195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,10240,0.06222755379147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,8192,0.05029955506324768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,7168,0.048410667313469775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,8192,0.02371999952528212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,8192,0.05307999915546841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,6144,0.042769776450263135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,7168,0.043755554490619235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,7168,0.021524444222450256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,7168,0.04717422193951077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,5120,0.03576622075504727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,6144,0.01901955571439531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,12288,0.07128000259399414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,6144,0.03897777862019009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,4096,0.030597332451078627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,5120,0.01701955497264862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,5120,0.03311289019054837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,10240,0.060247997442881264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,5120,0.037674665451049805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,3584,0.026888888743188646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,4096,0.02805066770977444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,4096,0.015223110715548197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,3072,0.023644444015291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,4096,0.03274044394493103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,3584,0.013933334085676404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,3584,0.02500533395343357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,3584,0.02996622191535102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,3072,0.01296622223324246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,3072,0.02221600049071842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,3072,0.02782755427890354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,2560,0.012081777883900536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,2048,0.016141333513789706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,2560,0.019596444235907663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,51200,768,0.02775466607676612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,2560,0.024454222785101995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,1536,0.01311377767059538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,2048,0.010969777901967367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,6144,0.04232888751559787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,2048,0.017144888639450073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,2048,0.021920000513394673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,1536,0.008914666871229807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,1536,0.01445777714252472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,768,0.008640888664457533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,1536,0.019582221905390423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,1024,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,1024,0.01200711147652732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,512,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,768,0.010295111272070143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,1024,0.016514667206340365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,768,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,256,0.004855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,2560,0.019865777757432725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,512,0.008716444174448649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,768,0.015062222878138224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,128,0.004217777815130022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,256,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,256,0.008616000413894653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,64,0.0038417776425679526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,256,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,128,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,32,0.0042204442951414315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,128,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,128,0.009723555710580613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,64,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,16384,1024,0.010086221827401055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,16384,32,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,65536,0.2953520086076525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,51200,0.24124622344970703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,65536,0.29356177647908527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,16384,0.0766577786869473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,65536,0.1302782164679633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,51200,0.2297004328833686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,512,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,51200,0.10373333427641128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,16384,0.03862844573126899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,16384,0.07731555567847358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,12288,0.05985689163208008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,512,0.013314666847387949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,10240,0.050273779365751475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,16384,0.08756444189283584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,12288,0.0303457776705424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,12288,0.05992533100975884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,8192,0.04194666610823738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,10240,0.05081244309743246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,10240,0.025823111335436504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,7168,0.03851288888189528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,10240,0.06083466609319051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,8192,0.021976000732845728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,8192,0.04238488939073351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,51200,0.2506222195095486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,65536,0.32028177049424916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,6144,0.03417244553565979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,8192,0.05052266518274943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,7168,0.020020445187886555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,7168,0.037464890215131975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,6144,0.01805688937505086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,6144,0.03299199872546726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,4096,0.02385422256257799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,6144,0.04153777824507819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,5120,0.016220443778567843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,5120,0.028282665544086035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,5120,0.0366888874106937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,3584,0.02103999919361538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,4096,0.014567111929257711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,4096,0.02385422256257799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,4096,0.03166044420666165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,3584,0.013560000393125745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,3072,0.018223999275101554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,3584,0.02141155633661482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,12288,0.06993510988023546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,2560,0.015432000160217285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,3584,0.028806222809685603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,3072,0.012553777959611682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,3072,0.019469334019554984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,2048,0.013040000365840064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,3072,0.026245334082179602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,2560,0.011640888949235281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,2560,0.017325333423084684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,5120,0.027671999401516382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,1536,0.010074666804737514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,2560,0.0240000006225374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,2048,0.009860444400045607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,7168,0.04651822315322029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,1024,0.008744888835483128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,2048,0.014911999305089315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,1536,0.008672888908121321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,2048,0.021519111262427434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,768,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,1536,0.01276622215906779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,1536,0.018977777825461496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,1024,0.010505778094132742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,768,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,1024,0.016480889585283067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,768,0.00906666699382994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,256,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,768,0.014112888111008538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,512,0.007954667011896769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,128,0.003866666720973121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,256,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,256,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,256,0.011878222227096558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,64,0.0037377778854635027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,128,0.006676444576846228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,128,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,12288,32,0.0038933331767717996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,128,0.009742221898502773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,64,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,65536,0.23526133431328666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,12288,32,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,1024,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,51200,0.2031057808134291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,65536,0.2496124373541938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,16384,0.060079110993279346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,65536,0.1265368858973185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,512,0.006979555719428592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,51200,0.19608444637722441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,512,0.013110222087966071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,12288,0.04863644308514065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,16384,0.06672355863783094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,16384,0.03702933258480496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,10240,0.04009244508213467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,16384,0.0869342221154107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,12288,0.05110133356518216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,12288,0.02888444397184584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,8192,0.03391644358634949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,12288,0.0682506693734063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,10240,0.024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,65536,0.3194577693939209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,51200,0.2478791077931722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,10240,0.05919022030300564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,7168,0.031073778867721558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,8192,0.03682577941152785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,8192,0.021312000022994146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,6144,0.025824889540672302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,8192,0.05000177688068814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,7168,0.019330667124854196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,7168,0.03245155678855048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,6144,0.028775112496482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,7168,0.04522755410936144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,6144,0.017595556047227647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,51200,0.09984444247351752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,6144,0.0410151117377811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,5120,0.015823110938072205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,5120,0.035651554663976036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,4096,0.02084622283776601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,3584,0.016184000505341422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,4096,0.030621333254708186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,10240,0.04369511206944784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,3584,0.013111111190583972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,3072,0.014416888356208801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,3584,0.028014222780863445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,3072,0.012073777616024017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,2560,0.012207110722859701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,5120,0.023599111371570166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,4096,0.018970666660202872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,2560,0.015404444601800708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,5120,0.02461333407296075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,2048,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,2560,0.01071911139620675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,2560,0.02348622183005015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,4096,0.014191110928853353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,1536,0.009024000002278222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,2048,0.009513777991135916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,2048,0.01350400017367469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,2048,0.021287999219364587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,1536,0.008054222497675154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,1024,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,3584,0.018928888771269057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,1536,0.011674666570292579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,768,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,1536,0.018743111027611625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,1024,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,1024,0.009189333352777693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,3072,0.017303999927308824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,1024,0.01481599940194024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,512,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,768,0.007251555720965068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,768,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,256,0.003839111162556542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,768,0.014127999544143677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,512,0.0069146665434042616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,128,0.0032151111712058387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,512,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,256,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,256,0.007058666812049017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,512,0.012652444342772165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,64,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,256,0.011531555818186866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,10240,32,0.003551111039188173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,128,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,128,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,128,0.009700444837411245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,65536,0.21237333615620932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,64,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,10240,32,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,51200,0.16131644778781468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,65536,0.20421955320570204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,16384,0.05650222301483154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,65536,0.07730666796366374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,51200,0.16035021675957575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,51200,0.06244888570573595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,3072,0.026076444321208533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,16384,0.055161777469846934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,16384,0.054655111498302884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,12288,0.01937333411640591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,12288,0.043410665459103055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,51200,0.14324533939361572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,10240,0.03753777676158481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,65536,0.18359200159708658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,8192,0.031522668070263334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,12288,0.04342844420009189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,10240,0.01698933376206292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,7168,0.027453333139419556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,8192,0.014984889162911309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,8192,0.030712889300452337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,10240,0.03819288810094198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,6144,0.02299733294381036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,8192,0.03264888789918687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,7168,0.013956444131003486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,7168,0.027232888672086928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,5120,0.020139555136362713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,7168,0.030273778571022883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,6144,0.012816888590653738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,6144,0.024068444967269897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,4096,0.01623288955953386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,6144,0.026960889498392742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,12288,0.041950222518708974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,5120,0.011724444727102915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,5120,0.020891555481486853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,3584,0.014677332507239448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,5120,0.02424977719783783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,4096,0.017909333109855652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,3072,0.013273778061072031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,4096,0.02168888847033183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,3584,0.009870222045315636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,3584,0.01651822196112739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,3584,0.020262221495310467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,3072,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,3072,0.015166223049163818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,10240,0.036743998527526855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,2048,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,3072,0.019023999571800232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,2560,0.01371911085314221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,2560,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,1536,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,2560,0.017672888106769986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,2048,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,2048,0.012044444680213928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,2048,0.016571554872724745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,1024,0.006220444622966979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,1536,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,1536,0.010017777482668558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,768,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,1536,0.014888889259762235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,1024,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,1024,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,4096,0.011038222246699862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,512,0.003951111187537511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,1024,0.012618666721714867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,768,0.006454222318198945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,768,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,256,0.0033893332713180115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,768,0.011365333365069495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,512,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,2560,0.011337777806652917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,512,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,512,0.011472889118724398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,256,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,256,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,64,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,256,0.01036800030204985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,128,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,32,0.003162666741344664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,128,0.008343111309740279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,64,0.006441777778996362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,65536,0.17652977837456596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,8192,32,0.006680000159475539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,16384,0.02458400030930837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,51200,0.13992355929480657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,65536,0.19963378376430937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,51200,0.15652533372243246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,16384,0.051522665553622775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,51200,0.06018666426340739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,65536,0.16950044367048475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,128,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,16384,0.053928888506359525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,12288,0.0403511126836141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,51200,0.13510578208499485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,16384,0.05227555500136482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,16384,0.02275288932853275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,10240,0.03218400147226121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,12288,0.01868266695075565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,12288,0.04191822144720289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,8192,0.0274311105410258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,12288,0.04253688785764906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,10240,0.03604888916015625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,10240,0.01651022169325087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,7168,0.023443554838498432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,10240,0.03753511110941569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,8192,0.014821334017647637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,8192,0.029927998781204224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,6144,0.020729778541458976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,8192,0.032144000132878624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,7168,0.013408888545301227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,7168,0.026642666922675237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,5120,0.017628444565667045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,7168,0.029141333368089464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,6144,0.012440889245933957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,6144,0.023569777607917786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,4096,0.01459022197458479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,6144,0.026214222113291424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,5120,0.011415999796655444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,5120,0.020601777566803824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,3584,0.013029333617952136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,5120,0.023489778240521748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,4096,0.010355555348926121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,4096,0.017610667480362784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,3072,0.011731555892361535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,4096,0.021152888735135395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,3584,0.009476444787449306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,3584,0.01623288955953386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,2560,0.010232888989978367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,3072,0.008832000195980072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,3072,0.014814222852389017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,2048,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,3072,0.01867911054028405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,2560,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,65536,0.07396533091862996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,2560,0.01329422245422999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,1536,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,2048,0.00740799970097012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,2048,0.011424889167149862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,1024,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,8192,128,0.0031635556370019913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,1536,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,1536,0.009682666924264696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,768,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,1536,0.013751111096805997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,1024,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,1024,0.008224888808197444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,512,0.0042204442951414315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,1024,0.011697777443461947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,768,0.006314666734801398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,768,0.007344889144102733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,256,0.00351200004418691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,768,0.011457777685589261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,512,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,512,0.006731555693679386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,128,0.003198222153716617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,3584,0.02011022302839491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,512,0.010504888991514841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,256,0.006498666687144174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,256,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,64,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,7168,32,0.0032479998966058097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,256,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,128,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,2560,0.01737866633468204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,128,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,64,0.0063946665161185795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,128,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,65536,0.1643973324033949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,7168,32,0.006505777852402792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,2048,0.015626667274369132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,51200,0.12969688574473062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,65536,0.16167199611663818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,16384,0.04412088791529337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,65536,0.0710008872879876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,51200,0.12508532736036512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,12288,0.03494133220778571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,16384,0.04430400000678169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,16384,0.05157155460781521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,51200,0.13139732678731283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,12288,0.034302221404181585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,12288,0.018067555295096505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,12288,0.04154577851295471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,10240,0.029253333806991577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,10240,0.01608088943693373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,8192,0.024780443973011438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,10240,0.03685422076119317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,8192,0.024672889047198828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,8192,0.014158222410413953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,7168,0.022056000100241765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,8192,0.03161066770553589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,51200,0.057518223921457924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,6144,0.0185253338681327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,16384,0.022133333815468684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,7168,0.01312444441848331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,7168,0.02223288847340478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,5120,0.015790222419632804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,7168,0.028423110644022625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,6144,0.012057777908113269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,6144,0.020048000746303134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,6144,0.026103999879625108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,10240,0.029191109869215224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,5120,0.011124444504578909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,3584,0.012070222033394707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,5120,0.023425777753194172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,4096,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,4096,0.01514311134815216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,3072,0.010861333045694562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,3584,0.008988444175985124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,3584,0.014127999544143677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,3584,0.01935822268327077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,3072,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,3072,0.01310400002532535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,2560,0.009619555539555019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,3072,0.018449778358141582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,2560,0.007780444290902879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,2560,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,2560,0.01724799970785777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,2048,0.009734222458468543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,65536,0.1643679936726888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,1536,0.006903999795516332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,2048,0.01514844430817498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,5120,0.017529777354664274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,1024,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,1536,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,1536,0.008574222524960836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,1536,0.01330488920211792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,768,0.004522666749027041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,4096,0.021055110626750525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,1024,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,1024,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,512,0.0038488888078265717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,1024,0.011750222080283694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,768,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,768,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,256,0.0032133333798911837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,768,0.011408000356621213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,512,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,2048,0.008301333420806462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,128,0.003121777748068174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,512,0.010446222292052375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,256,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,256,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,64,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,256,0.010317333042621613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,4096,0.01399466726515028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,128,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,6144,32,0.0029271110478374693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,128,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,65536,0.1431466605928209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,64,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,6144,32,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,51200,0.11240000194973415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,65536,0.15573067135281032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,65536,0.06844888793097602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,16384,0.041803555356131665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,51200,0.1228968832227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,51200,0.0552151103814443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,512,0.006260444306664997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,65536,0.16383111476898193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,12288,0.03230399886767069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,16384,0.020968889196713764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,16384,0.0432204438580407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,10240,0.02609955602222019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,51200,0.1304924488067627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,16384,0.05127733283572727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,12288,0.017518222332000732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,12288,0.033321777979532875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,8192,0.021990221407678392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,12288,0.04092622134420607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,10240,0.015797333584891427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,10240,0.028777778148651123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,7168,0.019069333871205647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,10240,0.03584533267550998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,8192,0.013807111316257052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,8192,0.024124445186720953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,6144,0.01646933290693495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,8192,0.030615998639000788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,7168,0.028025776147842407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,6144,0.0196088883611891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,5120,0.014363555444611443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,6144,0.011735999749766456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,6144,0.025449777642885845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,5120,0.010616888602574667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,4096,0.012068444656001197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,5120,0.02306933369901445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,2048,0.007401777638329401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,4096,0.014872888724009195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,3584,0.01092355532778634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,4096,0.020600888464185927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,7168,0.02197155521975623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,3584,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,3584,0.013791110780504016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,3584,0.019556444552209642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,3072,0.012829333543777466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,3072,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,3072,0.017992888887723286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,5120,0.01716088917520311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,2560,0.010966222319338055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,128,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,2048,0.00758844448460473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,2560,0.01683999929163191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,2560,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,1536,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,4096,0.009455111291673448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,2048,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,2048,0.009671111073758867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,1024,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,1536,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,7168,0.012748444245921241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,3072,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,1536,0.01349422252840466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,768,0.0038764443662431505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,1024,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,1024,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,512,0.0035075553589397003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,1024,0.011439111497667102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,768,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,2560,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,256,0.0028844444702068963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,768,0.010968888799349466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,512,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,512,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,128,0.002793777734041214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,512,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,256,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,256,0.005976000179847081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,64,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,256,0.009744000103738572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,5120,32,0.0026355555488003623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,128,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,128,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,128,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,65536,0.10582666926913792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,64,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,2048,0.01443733274936676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,1536,0.008307555483447181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,32,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,51200,0.08777778016196357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,65536,0.11784977383083767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,16384,0.03700800074471368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,65536,0.06612800227271186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,51200,0.05293777916166517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,51200,0.09283022085825603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,16384,0.020233778489960563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,5120,768,0.006683555742104848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,16384,0.03262844350602891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,65536,0.15579644838968912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,51200,0.12519200642903647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,10240,0.024359110328886244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,12288,0.01685244507259793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,12288,0.02568622264597151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,8192,0.020421332783169217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,10240,0.015187554889255099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,12288,0.03865244322352939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,10240,0.022481777601771887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,7168,0.017830222845077515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,10240,0.033935109774271645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,8192,0.013302221894264221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,6144,0.01514933341079288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,8192,0.019305777218606737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,8192,0.02946399980121189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,7168,0.012274666792816587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,7168,0.017494221528371174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,5120,0.013624000052611033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,7168,0.026699556244744196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,6144,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,6144,0.015916445189052157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,4096,0.010243555737866296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,6144,0.02437955637772878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,5120,0.009956444303194681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,12288,0.02956977817747328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,5120,0.02221333318286472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,4096,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,4096,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,16384,0.04830311073197258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,4096,0.0199626667631997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,3072,0.008264888491895463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,3584,0.011417778001891242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,3584,0.0181031111213896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,2560,0.0074684446056683855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,3072,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,3072,0.017095110482639737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,2560,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,2048,0.006551110910044775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,2560,0.015246222416559855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,5120,0.013996443814701505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,2048,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,1536,0.0053635554181204895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,2048,0.013810666898886362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,3584,0.00904888908068339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,1536,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,1024,0.00424177779091729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,3584,0.0083155557513237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,1536,0.012715555727481842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,3072,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,768,0.0038568890757030914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,1024,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,1024,0.006588444527652528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,1024,0.01181244436237547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,512,0.003208000005947219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,2560,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,768,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,768,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,256,0.0028888889484935333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,768,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,512,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,512,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,128,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,256,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,512,0.01070666644308302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,64,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,256,0.010039110978444418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,2048,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,128,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,4096,32,0.002814222127199173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,128,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,128,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,65536,0.0982453359497918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,1536,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,64,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,32,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,51200,0.07865955432256062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,65536,0.11706666151682536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,51200,0.09107999669180976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,51200,0.05186133252249824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,65536,0.15543822447458902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,51200,0.12410310904184978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,12288,0.028076443407270644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,16384,0.03199733297030131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,16384,0.01994400057527754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,10240,0.0233271105421914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,12288,0.016632889707883198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,16384,0.04764622118737963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,12288,0.02541333271397485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,8192,0.01923377811908722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,12288,0.03849688834614224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,4096,256,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,10240,0.022422222627533808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,7168,0.017076444294717576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,10240,0.033556444777382746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,8192,0.013072888884279462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,8192,0.019172444939613342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,6144,0.015112888481881885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,8192,0.029061334000693426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,7168,0.01742755538887448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,65536,0.0649706655078464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,7168,0.026584888497988384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,5120,0.013099555340078143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,6144,0.015829333000712924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,6144,0.02439466615517934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,5120,0.009677333136399588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,5120,0.013890667094124688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,5120,0.021760889225535925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,4096,0.011775111158688864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,3584,0.008771555291281806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,4096,0.01947022146648831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,10240,0.014921777778201632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,3584,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,3584,0.010793777803579966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,3584,0.0185199992524253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,7168,0.011759111450778114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,16384,0.032893333170149065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,3072,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,2560,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,3072,0.010063110954231685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,3072,0.016977777083714802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,4096,0.00963644435008367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,2048,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,2560,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,2560,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,2560,0.015434667468070984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,1536,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,2048,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,4096,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,2048,0.008303110798199972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,1024,0.004527111020353106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,2048,0.0140364451540841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,1536,0.006877333339717653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,1536,0.01277511070171992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,1024,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,3072,0.008264000217119852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,768,0.003847111016511917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,1024,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,512,0.003485333174467087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,1024,0.011370666325092316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,768,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,768,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,6144,0.010474666953086853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,768,0.01071111112833023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,512,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,512,0.010348444183667501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,128,0.0028115556471877625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,256,0.009292444421185387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,128,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,128,0.005869333528810077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,128,0.0076657773719893555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,1536,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,32,0.00286666676402092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,64,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,65536,0.09902844164106582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,32,0.005662222289376789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,256,0.002830222249031067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,65536,0.11112444930606419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3584,512,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,65536,0.06368533107969496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,51200,0.07277688715193006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,16384,0.030066665675905015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,65536,0.15390311347113714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,256,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,12288,0.021348444951905146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,51200,0.050973332590527005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,16384,0.01994044416480594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,51200,0.0872337751918369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,10240,0.0173680004146364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,16384,0.031001776456832886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,51200,0.12308622731102838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,12288,0.024824889169798955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,16384,0.047389331791136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,8192,0.015029332704014249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,12288,0.03793955511516995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,10240,0.014707555373509726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,10240,0.02164177762137519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,7168,0.013184888495339287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,8192,0.01276266657643848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,10240,0.03353066576851739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,8192,0.018601778480741713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,6144,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,8192,0.02829422222243415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,7168,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,7168,0.016975111431545682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,5120,0.010280888941552904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,7168,0.02649066680007511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,6144,0.010552888943089379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,6144,0.015494222442309061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,4096,0.008953777452309927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,6144,0.024041778511471216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,5120,0.009336888790130615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,5120,0.013419555293189155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,3584,0.008114666574531132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,5120,0.021515554851955835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,4096,0.008625778059164682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,4096,0.011081777513027191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,3072,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,4096,0.01891999940077464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,3584,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,3584,0.010310221877362993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,2560,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3584,64,0.002518222149875429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,12288,0.01644088824590047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,3072,0.009677333136399588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,2048,0.005762666463851929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,3072,0.015976000163290236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,2560,0.007620444728268518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,2560,0.008852444589138031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,1536,0.004961777892377642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,2560,0.015275556180212231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,2048,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,1024,0.0038986665507157645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,2048,0.008088889221350351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,2048,0.013802666631009845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,1536,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,768,0.003543111185232798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,1536,0.012552888856993781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,1024,0.006457777900828256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,1024,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,1024,0.011401777466138205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,512,0.0031511110977994073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,768,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,768,0.010469333165221743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,512,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,256,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,512,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,3584,0.018252443936136033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,256,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,128,0.00250133333934678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,256,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,1536,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,128,0.0053546664615472155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,64,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,128,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,128,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,3072,32,0.0024968888610601425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,768,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,65536,0.07972888814078437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,64,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,32,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,51200,0.06010044283337063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,65536,0.062407109472486705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,512,0.010444444086816577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,65536,0.10770666599273682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,3072,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,16384,0.02700177828470866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,65536,0.15379644764794245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,3072,256,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,51200,0.08574489090177749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,12288,0.02128177715672387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,16384,0.03052799900372823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,16384,0.01965777741538154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,51200,0.12308088938395183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,10240,0.01902044481701321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,16384,0.046689778566360474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,12288,0.016215999921162922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,12288,0.024320888850424025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,8192,0.014083556003040738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,10240,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,12288,0.03755111164516873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,10240,0.021192888418833416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,7168,0.012697777814335294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,8192,0.011704888608720569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,6144,0.011238222320874533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,8192,0.018350222044520907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,7168,0.011046222514576383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,8192,0.02817866537306044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,7168,0.016975111431545682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,5120,0.010313777459992303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,7168,0.02609866691960229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,6144,0.014836443795098199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,4096,0.008623999853928884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,6144,0.02385422256257799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,5120,0.009368889033794403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,5120,0.012465777496496836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,3584,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,5120,0.021419554948806763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,51200,0.04986133178075155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,4096,0.00868711123863856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,4096,0.01068800025516086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,3072,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,4096,0.018518222702874076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,3584,0.00831911133395301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,3584,0.010206222534179688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,2560,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,3584,0.017137777474191453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,3072,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,3072,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,2048,0.005881777654091517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,3072,0.015641777051819693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,2560,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,2560,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,1536,0.004524444540341695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,2560,0.015207111835479736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,2048,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,2048,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,1024,0.003863111138343811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,2048,0.013919111755159167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,1536,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,1536,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,768,0.003528888854715559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,6144,0.010096888575288985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,1536,0.01237777786122428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,1024,0.0064497776329517365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,1024,0.006471111128727595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,1024,0.011378666592968835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,768,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,768,0.005990222096443176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,256,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,768,0.011193777951929303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,512,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,128,0.0024862223201327855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,512,0.010057777994208867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,256,0.005702222386995952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,256,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,256,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,128,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,128,0.005285333428117964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,32,0.0024622221373849446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,128,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,65536,0.05823022127151489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,64,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,10240,0.03305066625277201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2560,32,0.005712000032265981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,51200,0.04617866542604235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,65536,0.06156355804867215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,65536,0.06948533323076037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,512,0.003162666741344664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,16384,0.020625778370433383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,65536,0.15310311317443848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,51200,0.05416711171468099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,51200,0.0483804444471995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,12288,0.01903555625014835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,512,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,16384,0.019298666053348117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,51200,0.12262133757273357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,16384,0.020988444487253826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,10240,0.015784889459609985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,16384,0.04622933268547058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2560,64,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,12288,0.015468445089128284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,8192,0.01330933305952284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,12288,0.03728888763321771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,10240,0.01309511148267322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,7168,0.011928000383906893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,10240,0.03305066625277201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,8192,0.011481777661376528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,8192,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,6144,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,8192,0.02816177739037408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,7168,0.0107351111041175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,7168,0.012125333150227865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,5120,0.009460444251696268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,7168,0.02588533361752828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,6144,0.010058666268984476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,6144,0.011053333679835001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,4096,0.008259555531872643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,6144,0.02351911034848955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,5120,0.009325332939624786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,5120,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,3584,0.007453333172533248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,5120,0.020566221740510728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,4096,0.008555555509196388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,4096,0.01809511085351308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,3072,0.006713777780532837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,3584,0.008307555483447181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,3584,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,12288,0.01703733371363746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,2560,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,3584,0.017335111896197002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,10240,0.015062222878138224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,3072,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,2048,0.004851555658711327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,3072,0.007698666718271043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,2560,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,2560,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,2048,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,2560,0.014870221416155497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,2048,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,1024,0.003524444583389494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,2048,0.01310222182008955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,1536,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,1536,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,768,0.003177777760558658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,1536,0.012135999898115793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,1024,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,1024,0.011074666347768573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,512,0.002789333255754577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,4096,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,768,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,768,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,256,0.002475555572244856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,768,0.010412444670995077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,512,0.005526222288608551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,128,0.002458666761716207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,512,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,256,0.005803555664088991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,256,0.009384888741705153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,3072,0.016075554821226332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,1536,0.004293333325121137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,128,0.004953777624501122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,32,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,128,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,65536,0.0496053331428104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,64,0.005197333378924264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,32,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,51200,0.03980088896221585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,1024,0.006534222099516127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,65536,0.06651733318964641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,65536,0.0600222216712104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,16384,0.01737333337465922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,65536,0.1537626716825697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,51200,0.047080000241597496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,51200,0.05235377616352505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,12288,0.014183110660976835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,16384,0.02015911042690277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,51200,0.12203466892242432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,2048,256,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,16384,0.04623022344377306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,10240,0.013034666577974955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,12288,0.016714667280515034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,128,0.005653333332803514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,12288,0.014210666219393412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,8192,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,12288,0.03692888882425096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,10240,0.01276711126168569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,10240,0.014779556128713818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,7168,0.0118568887313207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,10240,0.0326222214433882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,8192,0.011410666836632622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,8192,0.012081777883900536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,6144,0.010663111176755695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,8192,0.027479999595218237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,7168,0.010738666686746808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,7168,0.011706666813956367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,5120,0.009283555878533257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,6144,0.010055111514197456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,7168,0.02529777752028571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,6144,0.010460444622569613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,4096,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,6144,0.022600000103314716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,5120,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,5120,0.01993600030740102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,4096,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,3584,0.006907555378145642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,2048,64,0.0022053333620230355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,4096,0.018234666850831773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,3072,0.0063928887248039246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,3584,0.008053333395057255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,3584,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,3584,0.016512889001104567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,2560,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,3072,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,3072,0.007717333734035492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,2560,0.007323555648326874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,2048,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,2560,0.007182222273614671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,2560,0.014868444866604276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,2048,0.007177777588367462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,1536,0.004232888834344016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,16384,0.017851556340853374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,2048,0.013213333156373767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,1024,0.0035422220826148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,4096,0.008623999853928884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,1536,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,1536,0.012390221986505719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,768,0.003212444484233856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,1024,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,1024,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,1024,0.011376889215575324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,512,0.002903999967707528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,768,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,768,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,768,0.010400888820489248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,256,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,512,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,512,0.0052800000541739995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,512,0.010372444159454769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,128,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,256,0.005670222143332164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,256,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,2048,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,256,0.009487999810112847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,64,0.002248888835310936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1536,32,0.002475555572244856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,128,0.0057706667317284485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,128,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,65536,0.036632001399993896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,1536,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,128,0.008100444244013892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,51200,0.029699554045995075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,64,0.00509066672788726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1536,32,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,65536,0.057596445083618164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,16384,0.014125333891974555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,65536,0.06365955538219877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,51200,0.04630577895376417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,51200,0.05017333229382833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,65536,0.15258933438195124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,12288,0.011815999945004782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,51200,0.12144266234503852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,16384,0.01997066703107622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,10240,0.010604444477293225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,3072,0.015573332707087198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,16384,0.04550933175616794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,12288,0.014131555954615274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,12288,0.016037333342764113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,8192,0.012494222157531314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,12288,0.03673333260748122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,10240,0.012863111164834766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,7168,0.01143200033240848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,10240,0.013931555880440606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,5120,0.009335111412737105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,10240,0.03221333358022902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,8192,0.012074666718641916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,6144,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,8192,0.026755554808510676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,7168,0.010598222414652506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,7168,0.011212444139851464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,5120,0.00870222184393141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,7168,0.024510223004553053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,6144,0.009867555565304225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,6144,0.02221333318286472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,5120,0.009128889275921715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,4096,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,5120,0.009376000199053023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,5120,0.019783111082182992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,3584,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,4096,0.00851111114025116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,16384,0.016982222596804302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,4096,0.00832622249921163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,3072,0.005877333382765452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,4096,0.017504000001483493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,3584,0.008079110748238033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,3584,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,2560,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,3584,0.016784888174798753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,3072,0.007915555602974361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,3072,0.00741155528359943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,2048,0.004519999855094486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,8192,0.011328889264000786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,2560,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,2560,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,1536,0.0038657776183552216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,2560,0.014459555347760519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,2048,0.007044444481531779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,6144,0.01036711119943195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,2048,0.013716444373130798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,1024,0.0032382222513357797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,1536,0.006637333167923822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,1536,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,768,0.002871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,1536,0.012411555482281579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,1024,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,1024,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,512,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,1024,0.01108000013563368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,768,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,768,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,256,0.0023146667000320223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,512,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,768,0.010724444356229572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,512,0.005342222336265776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,128,0.0021137777302000257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,512,0.009723555710580613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,256,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,256,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,3072,0.015592000550693937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,64,0.0020186667227082783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,1024,32,0.0019528888579871918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,256,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,128,0.004977777600288391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,65536,0.03011200163099501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,128,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,2048,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,64,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,51200,0.024830222129821777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,1024,32,0.005019555489222209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,65536,0.0571075545416938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,65536,0.06239555279413859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,51200,0.04586844311820137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,51200,0.049160887797673546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,65536,0.15328089396158853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,12288,0.011335999601417117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,16384,0.019271110494931538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,16384,0.01680799987581041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,10240,0.010321777727868822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,16384,0.045356445842319064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,12288,0.014149333039919535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,12288,0.015397333436542086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,8192,0.00925244473748737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,12288,0.03640888796912299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,10240,0.013554666605260638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,10240,0.03124533428086175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,8192,0.011378666592968835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,7168,0.008708444734414419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,128,0.00565244464410676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,8192,0.026408000124825373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,6144,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,7168,0.010703999963071613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,7168,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,16384,0.011462222370836469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,7168,0.024264888630972967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,5120,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,6144,0.010212444596820408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,6144,0.0222746663623386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,4096,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,51200,0.12109243869781494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,5120,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,5120,0.00907911111911138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,3584,0.006305777778228124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,5120,0.01977066695690155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,4096,0.00853777759604984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,4096,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,3072,0.00562755556570159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,10240,0.01273155543539259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,3584,0.007807999849319458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,3584,0.016197333733240765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,8192,0.011838222543398539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,2560,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,3072,0.0075502221783002215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,3072,0.007699555820888943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,3072,0.015152888165579902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,2048,0.004580444345871608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,2560,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,2560,0.0069955554273393415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,2560,0.01479644411140018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,1536,0.0038888889054457345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,2048,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,2048,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,2048,0.013068444199032254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,1024,0.00332622230052948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,1536,0.00664444433318244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,1536,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,1536,0.012410666379663678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,768,0.003157333367400699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,1024,0.006370666540331311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,4096,0.01755555636352963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,1024,0.005746666755941179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,512,0.0028239999794297745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,1024,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,768,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,768,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,256,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,768,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,512,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,512,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,128,0.0021520000365045336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,512,0.009713778065310584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,256,0.005881777654091517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,256,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,64,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,256,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,128,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,6144,0.009913778139485253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,128,0.004986666556861666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,768,32,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,65536,0.023981332778930664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,128,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,51200,0.018765333626005385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,64,0.00500266667869356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,65536,0.05717155668470594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,768,32,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,65536,0.06076355775197347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,51200,0.04826755656136406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,51200,0.0449795557392968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,3584,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,12288,0.01240533341964086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,16384,0.016520889268981088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,51200,0.12088621987236871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,10240,0.011039999624093374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,16384,0.045239110787709556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,12288,0.013855111267831592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,12288,0.015107555521859063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,8192,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,12288,0.0355422231886122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,10240,0.012757333616415659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,10240,0.013449778159459433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,7168,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,10240,0.03091199861632453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,8192,0.011807999677128263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,8192,0.026105778084860906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,7168,0.011074666347768573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,6144,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,7168,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,7168,0.02401422295305464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,16384,0.010609777437316047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,5120,0.007344889144102733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,65536,0.15185066064198813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,6144,0.010231999887360467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,4096,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,6144,0.021903110875023737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,5120,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,5120,0.009368889033794403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,16384,0.018810666269726224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,3584,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,5120,0.019926221834288705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,4096,0.008621333373917473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,4096,0.008075555165608725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,3072,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,4096,0.01754044493039449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,3584,0.007711999946170383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,2560,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,3584,0.016491555505328707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,3072,0.007624000310897827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,3072,0.007375111182530721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,2048,0.0045422220395671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,3072,0.015455111861228943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,2560,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,2560,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,1536,0.0038311113086011675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,2560,0.01457333399189843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,2048,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,2048,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,1024,0.0031831111345026228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,2048,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,6144,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,1536,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,1536,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,768,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,1536,0.012294222083356647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,1024,0.005670222143332164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,1024,0.011017777853541903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,768,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,512,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,768,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,3584,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,768,0.01074488874938753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,256,0.0023226667609479693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,512,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,512,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,128,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,512,0.009664889011118147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,256,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,8192,0.01108977778090371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,256,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,256,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,128,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,32,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,65536,0.01681155628628201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,128,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,128,0.007620444728268518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,64,0.005125333451562458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,51200,0.014517333772447375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,512,32,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,65536,0.053391999668545194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,65536,0.059198220570882164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,1024,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,16384,0.009687110781669617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,65536,0.15120532777574328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,51200,0.04059111078580221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,12288,0.0083155557513237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,16384,0.01650399963061015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,51200,0.12091999583774143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,16384,0.018192888961897958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,10240,0.007383999725182851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,16384,0.043846223089430064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,12288,0.014101333088344999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,12288,0.014838222000333997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,8192,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,12288,0.034953776333067156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,10240,0.013066666821638743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,512,64,0.0018808889306253856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,7168,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,10240,0.030864887767367895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,8192,0.011676444775528379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,8192,0.011095999843544431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,6144,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,7168,0.010418666733635796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,8192,0.02605155607064565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,7168,0.010831111007266574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,5120,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,6144,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,6144,0.010013333625263637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,4096,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,6144,0.021928888228204515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,5120,0.009006222089131674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,5120,0.019552888141738046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,3584,0.006005333529578314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,4096,0.008364444805516137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,4096,0.01715288890732659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,3584,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,3072,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,10240,0.013426666458447775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,3072,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,3072,0.015117333994971381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,2560,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,3072,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,7168,0.02425688836309645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,5120,0.009047110875447592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,2560,0.007601777712504069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,2048,0.004229333251714706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,51200,0.04775466521581014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,2560,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,1536,0.0036151111125946045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,2560,0.014072888427310519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,2048,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,2048,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,1024,0.0031519999934567344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,2048,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,1536,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,1536,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,768,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,1536,0.011752888560295105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,3584,0.00797155582242542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,1024,0.006507555643717448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,1024,0.0058337777025169795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,512,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,768,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,1024,0.01071911139620675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,768,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,256,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,512,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,512,0.005402666827042897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,128,0.0018560000591807896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,512,0.01016088906261656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,256,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,256,0.004997333304749595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,256,0.009282666775915358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,64,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,128,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,128,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,256,32,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,65536,0.014486221803559197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,128,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,4096,0.008027555214034187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,64,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,51200,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,256,32,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,65536,0.04937511020236545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,65536,0.05857244465086195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,16384,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,3584,0.016160888804329764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,65536,0.1507893270916409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,51200,0.045915553967158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,51200,0.03997777899106344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,12288,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,51200,0.12019288539886475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,16384,0.016535111599498324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,16384,0.017959111266665988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,768,0.010055111514197456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,12288,0.014807111687130399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,16384,0.043856001562542386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,8192,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,12288,0.034965333011415266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,10240,0.0131217779384719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,10240,0.012888000243239932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,7168,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,10240,0.030525333351559106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,8192,0.011048888994587792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,8192,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,6144,0.005970666805903117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,8192,0.025811556312772963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,7168,0.01034577770365609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,7168,0.010974222587214576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,5120,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,7168,0.023694222172101338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,6144,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,6144,0.009743111001120673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,4096,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,6144,0.021686222818162706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,5120,0.009006222089131674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,5120,0.008981333010726506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,3584,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,5120,0.01960355540116628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,4096,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,10240,0.006205333189831839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,4096,0.016926222377353244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,3072,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,12288,0.014140443669425117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,3584,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,2560,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,3072,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,3584,0.016161777906947665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,2048,0.004853333450025982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,3072,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,2560,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,3072,0.015146666102939181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,2560,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,1536,0.003555555724435382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,2560,0.014458666245142618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,2048,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,2048,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,1024,0.0031679999083280563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,1536,0.0069013333155049225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,2048,0.013428444663683573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,1536,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,1536,0.012083555261294046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,1024,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,1024,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,512,0.0024862223201327855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,1024,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,768,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,768,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,4096,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,256,0.0021137777302000257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,768,0.01036000003417333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,512,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,512,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,128,0.0019493332753578823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,256,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,512,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,64,0.001783999924858411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,256,0.009722666607962714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,128,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,32,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,128,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,65536,0.011367999845080905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,128,0.007608888877762689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,51200,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,16384,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,64,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,12288,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,32,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,10240,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,65536,0.057096891933017306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,51200,0.04550844430923462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,16384,0.01797066628932953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,12288,0.014654222461912366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,7168,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,10240,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,6144,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,8192,0.011716444459226398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,5120,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,7168,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,4096,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,6144,0.009752000371615091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,3584,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,3584,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,128,256,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,5120,0.008978666530715095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,2560,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,4096,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,2048,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,3584,0.007424888511498769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,1536,0.003565333369705412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,2560,0.00693866651919153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,1024,0.0028684443483750024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,2048,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,768,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,128,768,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,512,0.002518222149875429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,1024,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,256,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,768,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,128,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,512,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,64,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,256,0.005200888961553574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,32,0.0017893332988023758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,128,0.0046746668716271715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,3072,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,65536,0.01629599928855896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,64,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,51200,0.013909333282046847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,16384,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,32,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,12288,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,3072,0.007197333706749811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,65536,0.05613955524232653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,51200,0.04531288809246487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,16384,0.017887999614079792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,8192,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,7168,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,64,8192,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,6144,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,5120,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,8192,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,4096,0.004557333472702238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,7168,0.01108977778090371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,3584,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,6144,0.010044444766309526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,3072,0.004853333450025982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,4096,0.007868444754017724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,2560,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,3584,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,2048,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,3072,0.0074106670088238185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,1536,0.003530666646030214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,2560,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,1024,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,2048,0.00664533343580034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,768,0.002510222295920054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,1536,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,512,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,1024,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,64,1536,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,256,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,768,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,128,0.0018008889423476325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,10240,0.01312622262371911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,64,0.0018151110659042995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,512,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,32,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,256,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,128,0.004974222017659081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,5120,0.008742222355471717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,64,0.0047377778424157035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,32,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,16384,0.4923866589864095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,16384,0.345399989022149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,256,32,10240,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,16384,0.1411866611904568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,256,32,12288,0.0147724449634552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,12288,0.3089759879642063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,10240,0.2538355456458198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,12288,0.25692976845635307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,8192,0.20374933878580728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,51200,1.080012427435981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,51200,1.5262426800198023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,10240,0.21191555923885771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,7168,0.17704444461398652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,10240,0.08995733658472697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,8192,0.17433777120378283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,8192,0.07370489173465304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,7168,0.06629244486490886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,6144,0.15178933408525255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,5120,0.13572443856133357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,6144,0.13406933678521052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,6144,0.05725333425733778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,7168,0.1512684424718221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,4096,0.10841421948538886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,3584,0.09261955817540486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,5120,0.04853777752982246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,4096,0.04110755523045858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,4096,0.09349333577685887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,12288,0.10558399889204238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,3072,0.08578488561842178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,3584,0.036440001593695745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,2560,0.06712710857391357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,3072,0.032906668053732976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,51200,0.41288444730970597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,2048,0.0592764417330424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,2560,0.028752889898088243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,2560,0.05968444214926826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,1536,0.04593777656555176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,2048,0.02542488939232296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,1024,0.0325288905037774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,1536,0.01978133287694719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,2048,0.05077777637375725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,768,0.02477866742346022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,1536,0.03965955641534593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,1024,0.015792888071801927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,1024,0.03159200151761373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,768,0.014188443620999655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,5120,0.10991466707653469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,256,0.01148622234662374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,768,0.025384000606007043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,512,0.013236444857385425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,512,0.021690666675567627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,128,0.00996888925631841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,256,0.012073777616024017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,64,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,3584,0.07920622163348727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,65536,128,0.01165511127975252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,32,0.008282666405042013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,3072,0.06845066944758098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,128,0.018230222993426852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,256,0.018964444597562153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,64,0.017887999614079792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,65536,32,0.017872888180944655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,65536,512,0.018087999688254464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,16384,0.3259493245018853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,16384,0.27254756291707355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,51200,0.8398844401041666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,65536,1.3203279707166884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,65536,1.1068702273898654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,51200,0.3332257800632053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,12288,0.24925687577989367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,16384,0.11367021666632758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,65536,0.4175111187828912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,12288,0.2015226682027181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,12288,0.08504533105426365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,8192,0.1587244404686822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,7168,0.1353208886252509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,10240,0.07170133458243476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,10240,0.19662133852640787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,8192,0.05760355790456136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,6144,0.11054222451315986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,8192,0.16074489222632513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,7168,0.11825866169399685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,6144,0.0449022220240699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,7168,0.051988446050220065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,6144,0.1070551143752204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,5120,0.09492266840404934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,5120,0.03930844532118903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,3584,0.07538933224148221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,4096,0.033746666378445096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,4096,0.0730115572611491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,3072,0.06637777884801228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,3584,0.03002133303218418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,10240,0.17456266615125868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,3584,0.0631271137131585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,2560,0.05425688955518934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,3072,0.027048889133665297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,2048,0.044165333112080894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,2560,0.023891554938422307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,2048,0.020407110452651978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,2560,0.04816177818510267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,3072,0.05548000004556444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,2048,0.04109511110517714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,1024,0.023334221707450017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,1536,0.01663822266790602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,5120,0.0930746661292182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,768,0.018379555808173288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,1024,0.013793778088357715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,1536,0.03291733397377862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,512,0.013728888498412238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,1024,0.02555822167131636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,51200,1.021444426642524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,256,0.010690666735172272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,768,0.012748444245921241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,4096,0.07775377564960055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,512,0.011687999798191918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,128,0.00904622260067198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,768,0.02067733307679494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,512,0.017869333426157635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,256,0.01034222212102678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,32,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,51200,128,0.010390222072601318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,64,0.014948444234000312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,32,0.015139554937680563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,128,0.01513066722287072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,65536,0.3962053457895915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,51200,0.3119235568576389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,65536,0.3627484374576145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,65536,0.14490755399068198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,16384,0.1143511136372884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,51200,0.2913724316491021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,64,0.007731555236710443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,12288,0.08384088675181071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,51200,0.11271022425757514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,16384,0.04114933477507697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,10240,0.07273599836561415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,12288,0.06992266575495402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,12288,0.03236444460021125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,51200,256,0.01610666679011451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,8192,0.05763644642300076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,10240,0.0594995551639133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,10240,0.02807377775510152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,7168,0.050511999262703784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,51200,1536,0.03151111139191522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,6144,0.04401866594950358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,8192,0.0499199993080563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,7168,0.021497777766651575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,5120,0.03801777627733018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,7168,0.04362755682733324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,6144,0.01891999940077464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,4096,0.03181333343187968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,5120,0.016978666186332703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,3584,0.02809600035349528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,5120,0.03311199943224589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,4096,0.014920888675583733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,3072,0.02479822271400028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,4096,0.027811557054519653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,3584,0.013835555149449242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,2560,0.0206631107462777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,16384,0.09185422129101223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,3072,0.012896888785892062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,3584,0.02513066596455044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,2048,0.016880000631014507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,3072,0.022062222162882488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,1536,0.013439111411571503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,2560,0.012039110892348819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,2560,0.019592000378502738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,2048,0.017097777790493436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,1024,0.01038044442733129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,1536,0.00887111077706019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,768,0.008651555412345463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,2048,0.011005333728260465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,1536,0.014319111903508505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,512,0.006836444553401735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,8192,0.023792889383104112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,1024,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,256,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,768,0.0075351107451650835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,1024,0.011787555283970304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,512,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,128,0.0039048890272776284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,768,0.010288000106811523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,64,0.003826666623353958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,256,0.0069875551594628235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,16384,32,0.0041751112374994485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,512,0.00902844468752543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,6144,0.03819555706448025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,16384,128,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,128,0.008276444342401292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,64,0.008275555239783393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,32,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,65536,0.2943413257598877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,51200,0.2376613352033827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,16384,0.07695821921030681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,65536,0.2920764552222358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,12288,0.05965422259436714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,65536,0.12820443842146131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,51200,0.22715377807617188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,10240,0.054435554477903575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,16384,0.03829244441456265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,16384,0.07689244217342801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,8192,0.04407199886110094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,12288,0.03000977635383606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,12288,0.05918222003512912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,10240,0.025967111190160114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,10240,0.0502017769548628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,6144,0.03298755486806234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,8192,0.042040887806150645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,7168,0.01995111174053616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,5120,0.027642667293548584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,6144,0.018087110585636564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,16384,256,0.008348444269763099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,6144,0.0329075555006663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,4096,0.02419288953145345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,5120,0.016176000237464905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,5120,0.02776088813940684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,4096,0.014484445254007975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,4096,0.023965333898862202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,3584,0.013415999710559845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,3072,0.018231999542978074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,3584,0.02125511070092519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,51200,0.10139822297626072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,2560,0.0151591118839052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,3072,0.012530666258600024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,7168,0.036877334117889404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,3072,0.019311999281247456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,2048,0.013083555632167392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,8192,0.021737777524524268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,2560,0.011535999675591787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,1536,0.01000711073478063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,2048,0.009927111367384592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,7168,0.037010666396882795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,1024,0.008948444492287105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,1536,0.00834222220712238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,2048,0.01478488908873664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,768,0.0075457774930530125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,1536,0.012744888663291931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,512,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,1024,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,768,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,1024,0.010044444766309526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,768,0.009031111167536842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,512,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,3584,0.020396444532606337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,128,0.0034933334423436057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,512,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,256,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,64,0.003168888803985384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,256,0.007527111305130853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,32,0.003232888877391815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,128,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,64,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,65536,0.23795021904839408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,32,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,12288,2560,0.01720711092154185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,51200,0.18370489279429117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,65536,0.24898134337531197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,16384,0.06212800078921848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,65536,0.12373422251807319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,12288,0.05258399910397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,51200,0.09937689039442275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,12288,256,0.004564444637960858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,16384,0.06566222508748372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,10240,0.043803556097878345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,16384,0.03665511144532098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,12288,0.028620445066028174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,8192,0.03555466731389364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,12288,0.050755557086732656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,7168,0.031397332747777305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,10240,0.024784889486100938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,8192,0.021311110920376245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,12288,128,0.006800000038411882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,6144,0.026529777381155226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,10240,0.043228444125917226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,5120,0.022704000274340313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,8192,0.03645688957638211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,4096,0.01904355486234029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,6144,0.01759288873937395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,5120,0.015844444433848064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,6144,0.028059555424584284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,3584,0.016401777664820354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,5120,0.02427644365363651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,4096,0.014144889182514615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,3072,0.014544000228246054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,4096,0.020630222227838304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,3584,0.013076444466908773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,2560,0.012511110968059964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,3584,0.01885333326127794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,3072,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,2048,0.010952889091438718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,3072,0.01719555589887831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,51200,0.19405244456397164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,1536,0.008778666456540426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,2560,0.01534400052494473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,1024,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,1536,0.00812266684240765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,1536,0.011418666276666852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,768,0.005623111294375525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,1024,0.007339555356237624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,1024,0.009290667043791877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,512,0.0047155556579430895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,7168,0.019241778386963736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,768,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,256,0.0038782221575578055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,768,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,512,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,128,0.0032106666929192017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,256,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,256,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,64,0.0031840000301599503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,128,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,10240,32,0.003237333355678452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,128,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,2560,0.010468444062603844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,64,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,10240,2048,0.008851555486520132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,32,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,65536,0.24311555756462944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,2048,0.013441777891582914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,51200,0.1971671051449246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,65536,0.07737333244747586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,65536,0.20104355282253691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,16384,0.06328977478875054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,51200,0.15814133485158285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,12288,0.04919377631611294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,51200,0.06203555398517185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,7168,0.03230133321550157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,16384,0.023707555400000677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,16384,0.054415110084745616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,12288,0.01924977699915568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,10240,512,0.007625777688291337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,8192,0.03510311245918274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,12288,0.04236800140804715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,10240,0.016931555337376065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,7168,0.030232889784706965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,10240,0.03639644384384155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,8192,0.014839111102951897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,6144,0.026505778233210247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,7168,0.013798221945762634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,5120,0.02276177704334259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,8192,0.030239999294281006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,4096,0.0191920002301534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,7168,0.02695377833313412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,6144,0.012729778057999082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,5120,0.011953777737087674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,6144,0.02389066749148899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,4096,0.010672888822025724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,5120,0.020611556039916146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,3072,0.014754666222466363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,3584,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,4096,0.017829333742459614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,2560,0.012662221988042196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,3584,0.016255110502243042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,3072,0.008742222355471717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,2048,0.01074488874938753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,3072,0.014947555131382413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,2560,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,2560,0.013516444298956128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,2048,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,10240,0.042861332496007286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,1024,0.007107555866241455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,2048,0.012048000262843238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,1536,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,768,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,1536,0.00979733301533593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,512,0.004179555508825514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,1024,0.006656888872385025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,768,0.006567111031876669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,1024,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,256,0.0034924443397257063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,768,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,512,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,128,0.0031679999083280563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,512,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,256,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,64,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,32,0.0032275555034478507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,256,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,128,0.0064293332397937775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,64,0.0064204442832205035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,65536,0.18136889404720732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,8192,32,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,51200,0.14212621582878962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,65536,0.07283466392093234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,65536,0.19814133644104004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,16384,0.04877510997984144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,12288,0.03813599877887302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,51200,0.15425866179996067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,51200,0.05886933538648817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,16384,0.02329155637158288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,10240,0.03403110967742072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,16384,0.05328088998794556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,12288,0.018564444449212816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,8192,0.027848889430363972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,10240,0.016528000434239704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,10240,0.03540444374084473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,8192,0.014543111125628153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,8192,128,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,6144,0.020744888318909537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,7168,0.013457777599493662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,7168,0.02607733342382643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,5120,0.017575111654069688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,6144,0.012416889270146688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,6144,0.023345778385798138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,4096,0.014923555983437432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,5120,0.011416888899273343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,1536,0.008800889054934183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,3584,0.013075555364290873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,5120,0.020212444994184706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,4096,0.009829333258999718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,3072,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,4096,0.01754666699303521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,3584,0.009158222211731805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,3584,0.016192888220151264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,2560,0.01070666644308302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,3072,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,2048,0.008949333594905006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,3072,0.014539556370841132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,12288,0.04121333360671997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,2560,0.008112000094519721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,7168,0.024822221861945257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,1536,0.0079297779334916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,2048,0.007246221933099959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,8192,0.02927910950448778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,2048,0.011342222491900126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,1024,0.0063786668082078295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,1536,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,768,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,1536,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,1024,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,512,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,768,0.00630044440428416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,1024,0.008068444828192392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,768,0.007463999920421177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,512,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,128,0.002954666606254048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,512,0.006911110960774952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,256,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,256,0.006606222026877933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,64,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,32,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,7168,128,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,128,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,65536,0.17002222273084852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,64,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,32,0.006497777584526274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,51200,0.13598844740125868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,65536,0.15913332833184138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,7168,2560,0.01313866674900055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,65536,0.07036266724268596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,16384,0.04803911182615492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,12288,0.0369742214679718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,51200,0.12346755133734809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,51200,0.0573671129014757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,16384,0.02179644505182902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,10240,0.03147644466824002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,7168,256,0.003548444559176763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,16384,0.043614221943749316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,12288,0.017982222967677645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,8192,0.02546577817863888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,12288,0.033739556868871055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,10240,0.01592088904645708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,7168,0.0224906669722663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,6144,0.01959288948112064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,10240,0.02892711096339756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,8192,0.014306666122542487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,5120,0.016674665941132438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,7168,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,8192,0.024506666594081458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,6144,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,7168,0.021852445271280076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,5120,0.011008000208271874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,6144,0.019723556107944913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,3584,0.013302221894264221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,4096,0.009391110804345872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,5120,0.017135999268955655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,3072,0.011373333632946014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,4096,0.015096889601813423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,3584,0.008967999782827165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,2560,0.009635555247465769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,3584,0.013797333670987023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,3072,0.008329778081840938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,8192,3584,0.016591999265882704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,2048,0.008401778009202745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,3072,0.012812444733248817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,2560,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,1536,0.006731555693679386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,1024,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,2048,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,1536,0.006917333437336816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,2048,0.009739555418491364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,768,0.004228444563017951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,512,0.0035377778112888336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,1536,0.00846400029129452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,1024,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,1024,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,768,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,512,0.006044444524579578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,768,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,128,0.0026684444811609057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,256,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,512,0.006526222245560751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,64,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,32,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,256,0.006119110931952794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,128,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,65536,0.1504408915837606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,64,0.006099555641412735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,32,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,51200,0.12496266100141738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,65536,0.15334399541219076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,65536,0.06822755601671007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,16384,0.04253955682118734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,6144,2560,0.011511999699804517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,51200,0.12064888742234971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,12288,0.033189333147472806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,16384,0.02106488909986284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,10240,0.027777777777777776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,16384,0.04259555538495382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,12288,0.017500445246696472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,256,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,8192,0.02358755634890662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,12288,0.032771554258134626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,10240,0.015511110424995422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,7168,0.02022133270899455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,10240,0.027759111589855615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,8192,0.013811555173661975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,6144,4096,0.013739555246300168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,6144,0.018287110659811232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,8192,0.024077332682079736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,5120,0.01605866683853997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,7168,0.012780444489585029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,6144,0.011725333001878528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,7168,0.021456888980335657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,4096,0.01348622226052814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,6144,0.01921688848071628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,5120,0.010372444159454769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,3584,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,5120,0.01682133310370975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,4096,0.00902844468752543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,3072,0.01069777790043089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,4096,0.014891554911931356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,3584,0.008482666479216682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,2560,0.009927999642160203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,51200,0.054775999652014844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,3072,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,2048,0.008249777886602614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,3072,0.012458666331238218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,2560,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,1536,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,2560,0.010975999964608086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,1024,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,1536,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,2048,0.009324444664849175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,768,0.003940444439649582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,1536,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,1024,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,6144,128,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,512,0.00351555562681622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,1024,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,256,0.0028817777832349143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,768,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,512,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,128,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,256,0.005879111174080107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,512,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,64,0.002502222235004107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,256,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,5120,32,0.00282222218811512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,128,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,128,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,64,0.005984000033802456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,3584,0.013639111485746173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,5120,32,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,51200,0.09690666861004299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,2048,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,65536,0.11606044239468044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,16384,0.03871733281347487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,65536,0.06560444169574313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,12288,0.029125332832336426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,51200,0.08974488576253255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,51200,0.052448888619740806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,10240,0.024333332975705464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,12288,0.01682133310370975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,5120,768,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,8192,0.020100444555282593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,10240,0.015115555789735583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,10240,0.021985777550273474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,7168,0.017785777648289997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,8192,0.013343111508422427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,6144,0.015457777513398064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,8192,0.019080888893869188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,7168,0.012154666913880242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,5120,0.013400889105266996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,7168,0.01722311145729489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,6144,0.01073155552148819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,4096,0.012275555895434486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,6144,0.01551999979548984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,65536,0.12147733900282119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,5120,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,3584,0.010478222535716163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,5120,0.013798221945762634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,4096,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,3072,0.009301332963837517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,4096,0.01236533373594284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,3584,0.008344888687133789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,2560,0.008615111311276754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,3584,0.011006222003036074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,3072,0.00793599999613232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,2048,0.006832888970772426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,3072,0.009951999617947472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,2560,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,1536,0.0057751110030545135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,16384,0.032272001107533775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,2048,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,2560,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,12288,0.025383111503389146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,1536,0.006908444480763541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,2048,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,1024,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,1024,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,512,0.003489777859714296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,768,0.006188444379303191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,768,0.006254222244024277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,256,0.0029013332807355454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,512,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,512,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,128,0.0027955555253558685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,256,0.0058746664888328975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,256,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,64,0.00253866674999396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,128,0.005697777701748743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,32,0.0028000000036425064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,128,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,4096,16384,0.02032177812523312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,64,0.00562755556570159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,65536,0.09698666466606988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,1024,0.004596444467703502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,32,0.005663110978073544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,51200,0.07885689205593534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,4096,768,0.004045333299371931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,16384,0.031925333870781794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,4096,1536,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,65536,0.11566310458713108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,51200,0.05121244324578179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,12288,0.028352889749738906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,10240,0.02310488952530755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,16384,0.019928000039524503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,16384,0.03163733416133457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,12288,0.016551999582184684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,8192,0.019111999207072787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,10240,0.015042665931913586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,12288,0.025401777691311304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,7168,0.01698755555682712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,10240,0.02200177808602651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,6144,0.015046222342385186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,8192,0.012915555801656512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,5120,0.013077333569526672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,7168,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,8192,0.018915555543369718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,4096,0.010480889015727572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,6144,0.010440888504187265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,7168,0.017103110750516255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,3584,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,5120,0.009390222529570261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,4096,0.00869777798652649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,5120,0.013789333403110504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,3072,0.008526222573386298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,4096,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,3584,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,2560,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,3584,0.01055111073785358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,3072,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,2048,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,3072,0.009711110757456886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,1536,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,65536,0.06434577703475952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,2560,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,2048,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,1024,0.004845333182149463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,2048,0.008103999826643202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,51200,0.08931910991668701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,1536,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,1024,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,1536,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,512,0.003568888952334722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,768,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,1024,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,256,0.00259911103381051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,768,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,512,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,128,0.0024862223201327855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,512,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,256,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,64,0.002236444503068924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,256,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,32,0.002371555608179834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,128,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,6144,0.015511110424995422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,128,0.005492444253630108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,65536,0.08321333593792386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,51200,0.07237066825230916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,32,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,65536,0.10922577646043564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,16384,0.02833066715134515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3584,2560,0.007592889169851939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,51200,0.05018577641910977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,51200,0.08510133292939927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,12288,0.025958221819665697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,10240,0.018033777674039204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,16384,0.019858666592174105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,12288,0.016399110356966656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3584,768,0.004184000194072723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,8192,0.014743111199802823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,12288,0.024540444215138752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,10240,0.014489778214030795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,7168,0.01333866682317522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,10240,0.02134399943881565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,8192,0.012462221913867526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,6144,0.011847111086050669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,7168,0.01105955574247572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,7168,0.016725333200560678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,5120,0.010168888502650792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,6144,0.010105777945783403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,6144,0.015264888604482016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,4096,0.009082666701740688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,5120,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3584,64,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,5120,0.013305777476893531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,65536,0.06338755289713542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,4096,0.008650666309727563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,3072,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,4096,0.011431111229790581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,3584,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,2560,0.006779555645253923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,3584,0.010064000056849586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,3072,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,2048,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,16384,0.030597332451078627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,3072,0.009631111390060848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,2560,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,2560,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,2048,0.007320000065697565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,1024,0.0037137779096762338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,2048,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,1536,0.006688000013430913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,8192,0.01845066746075948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,768,0.0032257777121331957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,1536,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,512,0.0028284444577164124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,1024,0.006380444599522485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,768,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,1024,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,256,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,768,0.006223111102978389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,512,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,128,0.0022444443570242985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,512,0.005768888940413793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,3584,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,128,0.005882666756709416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,256,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,32,0.0021946666141351065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,128,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,65536,0.08112977610694037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,64,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,3072,32,0.005632888939645555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,51200,0.06537866592407227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,1536,0.004568888909286923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,65536,0.10684355762269761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,65536,0.062149332629309766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,16384,0.02664799988269806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,12288,0.020822222034136455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,16384,0.01959111127588484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,51200,0.08383999930487739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,16384,0.0299288895395067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,10240,0.021984888447655573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,12288,0.016201777590645682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,8192,0.01463822192615933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,12288,0.024145777026812237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,10240,0.01425066590309143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,8192,0.011706666813956367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,10240,0.02074044446150462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,3072,256,0.005722666780153911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,8192,0.018068444397714403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,3072,64,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,6144,0.011042666931947073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,5120,0.009917333722114563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,7168,0.010995555255148145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,7168,0.016265778077973258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,4096,0.00906044410334693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,6144,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,3584,0.008618666893906063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,6144,0.014440889159838358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,4096,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,3072,0.008105778031879002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,4096,0.010880889164076911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,3584,0.008299555215570662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,2560,0.007144889069928064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,3584,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,3072,0.007879999776681265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,2048,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,3072,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,2560,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,1536,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,2560,0.008631111019187504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,2048,0.007199111084143321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,1024,0.0041893331540955436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,2048,0.008016889293988546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,1536,0.006738666858938005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,7168,0.012724444270133972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,768,0.0038559999730851916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,1536,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,512,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,1024,0.006515555497672822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,1024,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,768,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,256,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,768,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,256,0.005858666780922148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,5120,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,128,0.00254844439526399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,256,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,5120,0.012645333177513547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,64,0.0021840000732077495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2560,32,0.0021964444054497615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,128,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,65536,0.0640995568699307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,64,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,32,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,51200,0.04958044489224752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,51200,0.055123554335700146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,65536,0.06064089139302572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,16384,0.022537777821222942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,51200,0.04782933327886793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,16384,0.019303111566437613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,16384,0.020258666740523446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,12288,0.018160000443458557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2560,512,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,10240,0.017438222964604694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,512,0.005664000080691443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,12288,0.015446222490734525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,8192,0.013407111167907715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,10240,0.013633777697881063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,12288,0.016639111770523917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,7168,0.012110222544935016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,10240,0.015018666783968607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2560,128,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,8192,0.011424000064531961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,6144,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,8192,0.013036444783210754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,7168,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,5120,0.00958577791849772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,7168,0.011798222031858234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,6144,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,6144,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,65536,0.06724800003899468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,3584,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,5120,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,51200,0.05275466707017687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,4096,0.008307555483447181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,3072,0.006805333412355847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,3584,0.00832177781396442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,2560,0.006140444427728653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,3072,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,2048,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,3072,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,2560,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,1536,0.004482666651407878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,2560,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,2048,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,1024,0.00351200004418691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,2048,0.006759111252095964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,1536,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,1536,0.006305777778228124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,768,0.003168888803985384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,1024,0.006518222391605377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,512,0.00282222218811512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,1024,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,768,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,256,0.002505777817633417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,5120,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,512,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,512,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,128,0.002453333387772242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,256,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,256,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,64,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,32,0.0025333333760499954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,128,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,3584,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,65536,0.048729777336120605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,128,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,64,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,51200,0.03972266780005561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,32,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,16384,0.01717422240310245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,65536,0.05852355559666952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,65536,0.06389511293835111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,51200,0.0473911099963718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,12288,0.01462044484085507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,51200,0.05058311091528999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,16384,0.01843377782238854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,16384,0.019900444481107924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,12288,0.014186667071448432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,8192,0.01275644451379776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,12288,0.016709332664807636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,2048,4096,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,10240,0.014310222533014087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,2048,768,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,8192,0.011378666592968835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,6144,0.010443555812040964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,8192,0.012187555432319641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,2048,4096,0.008604444563388824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,7168,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,5120,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,7168,0.011239111423492432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,4096,0.007820444802443186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,6144,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,5120,0.009318222602208456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,6144,0.01035199976629681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,5120,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,3584,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,4096,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,4096,0.00830488900343577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,3072,0.006562666760550604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,3584,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,2560,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,3584,0.008003555238246918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,3072,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,2048,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,3072,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,10240,0.01551822159025404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,2560,0.007592000067234039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,1536,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,2560,0.007000000112586551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,1024,0.0037928890022966596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,2048,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,10240,0.013117333253224691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,1536,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,7168,0.011751111182901593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,1024,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,1536,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,512,0.0029973333908451926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,768,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,1024,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,256,0.002623999905255106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,768,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,512,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,128,0.0021671110557185277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,256,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,512,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,64,0.0018800000349680583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,32,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,256,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1536,128,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,65536,0.03561955690383911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,128,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,51200,0.02927466564708286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,64,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,32,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,16384,0.014073777529928418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,65536,0.05772622426350912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,65536,0.061496886942121715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,51200,0.04629066586494446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,12288,0.012060444388124677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,10240,0.011015111373530494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,16384,0.01683200067943997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,12288,0.014135999812020196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,16384,0.019270223047998216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1536,2048,0.006649777707126405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,8192,0.009902222288979424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,12288,0.015507555670208402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,7168,0.009832888841629028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,10240,0.013790222505728403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,6144,0.00906666699382994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,7168,0.010701333483060202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,7168,0.011015111373530494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,5120,0.0086986662613021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,6144,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,6144,0.010056888891590966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,4096,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,5120,0.009290667043791877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,5120,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,3584,0.0063893331421746155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,4096,0.008591111335489485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,4096,0.008199111455016667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,3072,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,3584,0.00817599975400501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,3584,0.0079297779334916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,51200,0.04854844344986809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,2560,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,3072,0.007794666621420119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,2048,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1536,768,0.0035004446076022256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,2560,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,1536,0.00388355553150177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,2560,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,8192,0.01130222198035982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,2048,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,1024,0.0031742221779293488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,8192,0.01201777739657296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,2048,0.00666933341158761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,768,0.0028844444702068963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,1536,0.006650666809744305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,512,0.002521777732504739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,1536,0.006216889040337668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,1024,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,256,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,1024,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,512,0.0058702222175068325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,768,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,128,0.0021280000607172647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,512,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,256,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,64,0.001975111042459806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,1024,32,0.0021199999998013177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,128,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,65536,0.0302800006336636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,128,0.00508711114525795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,64,0.004954666727119022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,32,0.004970666848950916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,51200,0.024225778049892847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,3072,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,65536,0.05711911122004191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,65536,0.06016000111897787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,10240,0.012741333080662621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,16384,0.012040889097584618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,12288,0.011551111108726926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,51200,0.04579377836651272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,16384,0.016724444097942777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,51200,0.047535998953713306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,10240,0.010503110786279043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,16384,0.018340445227093168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,12288,0.013927999469969006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,12288,0.01512977812025282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,10240,0.012744888663291931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,1024,768,0.005974222388532426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,7168,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,10240,0.013411555853154926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,8192,0.011141333315107556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,6144,0.008961777720186446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,8192,0.011761777930789523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,5120,0.007860444486141205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,1024,256,0.005196444276306364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,6144,0.009822222093741098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,4096,0.006897777732875612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,6144,0.010001777774757808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,5120,0.00928177767329746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,3584,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,5120,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,4096,0.008301333420806462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,3072,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,3584,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,3584,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,2560,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,3072,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,2048,0.004554666578769684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,2560,0.0074240002367231585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,3072,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,1536,0.0038986665507157645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,8192,0.009560888840092553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,2560,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,1024,0.0034346667428811393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,2048,0.007257777783605788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,2048,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,1536,0.006670222100284364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,768,0.0031511110977994073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,1024,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,1536,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,7168,0.010695999695195092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,1024,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,7168,0.010911111202504901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,768,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,256,0.0024666666156715816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,128,0.0021564445147911706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,768,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,256,0.005640888793600931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,64,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,128,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,4096,0.008350222474998897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,32,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,128,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,65536,0.024880000286632117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,64,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,32,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,65536,0.056851552592383496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,65536,0.05871733029683431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,16384,0.01070577816830741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,51200,0.04485955503251818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,51200,0.046725332736968994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,12288,0.009099555512269339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,16384,0.016792888442675274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,16384,0.01787911189926995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,10240,0.011101333631409539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,768,512,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,12288,0.014172444740931192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,8192,0.009509333305888707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,12288,0.014871110518773397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,768,512,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,10240,0.012881778180599213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,7168,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,512,0.0052879999081293745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,10240,0.013221333424250284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,6144,0.007958222594526079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,8192,0.011094222466150919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,768,256,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,5120,0.007432888779375289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,8192,0.011732444167137146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,7168,0.010673777924643623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,4096,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,6144,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,3584,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,5120,0.009322666459613377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,6144,0.010079111489984725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,3072,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,5120,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,4096,0.008023999631404877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,3584,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,2560,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,3584,0.007723555796676212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,3072,0.007644444704055786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,2048,0.004736888739797804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,3072,0.0073831114504072405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,2560,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,2560,0.0069439998931354946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,1536,0.004187555362780889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,2048,0.0069288888739215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,1024,0.0032133333798911837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,2048,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,1536,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,768,0.0028880000528362063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,1536,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,1024,0.006260444306664997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,512,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,1024,0.005714666512277391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,768,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,768,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,256,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,512,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,128,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,7168,0.010823110739390055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,512,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,256,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,4096,0.008504889077610439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,256,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,512,128,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,51200,0.022650667362742957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,32,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,128,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,65536,0.018276444739765592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,51200,0.015746666325463187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,64,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,65536,0.05337599913279215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,16384,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,512,32,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,51200,0.04088000125355191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,12288,0.009135999613338048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,51200,0.046238223711649575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,16384,0.016541333662139047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,10240,0.00849599970711602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,12288,0.01391377713945177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,8192,0.0075937774446275495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,12288,0.014577777849303352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,10240,0.013403555585278405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,7168,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,8192,0.011076444553004371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,8192,0.011372444530328115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,6144,0.006691555596060223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,7168,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,512,64,0.002056888822052214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,7168,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,5120,0.008005333443482717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,6144,0.009697777529557547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,4096,0.006790222393141852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,6144,0.00980266680320104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,5120,0.0090524446633127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,3584,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,5120,0.008964444200197855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,3072,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,4096,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,3584,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,65536,0.05766310956743029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,2560,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,3584,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,3072,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,2048,0.00424799985355801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,16384,0.017535110314687092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,2560,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,1536,0.00369777778784434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,2560,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,2048,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,1536,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,2048,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,768,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,1536,0.005971555494599872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,1024,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,1024,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,768,0.005979555762476391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,768,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,256,0.0021608887861172357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,512,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,4096,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,128,0.0021235555824306277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,256,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,128,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,64,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,128,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,3072,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,10240,0.012952889005343119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,1024,0.0031866667171319327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,64,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,51200,0.012658666405412885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,256,32,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,65536,0.049288001325395375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,65536,0.05630222294065687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,256,512,0.002529777793420686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,51200,0.039721777041753135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,51200,0.04440711273087395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,512,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,16384,0.016520000166363187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,12288,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,10240,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,16384,0.017588444881969027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,12288,0.013803555733627744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,8192,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,10240,0.01275644451379776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,256,256,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,12288,0.014471999473041959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,7168,0.00795555528667238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,10240,0.012855999999576144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,6144,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,8192,0.011114666859308878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,7168,0.010390222072601318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,5120,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,7168,0.010473777850468954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,6144,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,65536,0.014801777071423002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,6144,0.0096133334769143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,5120,0.009016888837019602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,3584,0.005962666538026597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,5120,0.008634666601816813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,4096,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,3072,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,4096,0.007735999921957652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,16384,0.01036977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,3584,0.007968000239796108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,3584,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,3072,0.007603555917739868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,2048,0.0041955556306574075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,3072,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,2560,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,1536,0.0035439998739295532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,2048,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,2048,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,1024,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,1536,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,1536,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,768,0.0028684443483750024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,1024,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,8192,0.01127466642194324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,512,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,1024,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,4096,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,768,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,256,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,768,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,128,0.0018826667219400406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,512,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,256,0.005881777654091517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,512,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,64,0.001793777777089013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,225,128,128,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,256,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,2560,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,65536,0.017301332619455125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,128,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,51200,0.014799111419253878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,16384,0.00811288919713762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,64,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,32,0.004874666531880696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,12288,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,10240,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,128,2560,0.006637333167923822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,8192,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,65536,0.05491111013624403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,7168,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,51200,0.04387199878692627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,6144,0.005753777921199799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,16384,0.01739555597305298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,5120,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,4096,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,12288,0.014780445231331719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,10240,0.012786666552225748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,3584,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,8192,0.011362666885058085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,3072,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,7168,0.01071111112833023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,2560,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,2048,0.004209777961174647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,5120,0.008623110751310984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,1536,0.003546666767862108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,3584,0.007384888827800751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,3072,0.007057777709431118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,1024,0.0028488888508743713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,2560,0.0069022224181228215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,768,0.002847111059559716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,512,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,2048,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,256,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,1536,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,128,0.0017866666118303935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,128,32,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,64,0.0017973333597183228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,1024,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,64,32,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,768,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,512,0.005187555733654234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,65536,0.017613333132531907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,256,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,51200,0.015134221977657743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,64,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,32,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,16384,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,12288,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,65536,0.054746667544047035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,10240,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,51200,0.043918222188949585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,8192,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,16384,0.017295110556814406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,6144,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,7168,0.006020444548792309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,12288,0.014615999327765571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,6144,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,5120,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,4096,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,10240,0.01311644415060679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,4096,0.006559111177921295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,3584,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,8192,0.011448889142937131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,3072,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,6144,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,2560,0.004849777867396672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,4096,0.008089777496125963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,2048,0.004224888980388641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,3584,0.00740444411834081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,1536,0.0034888887570963967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,3072,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,1024,0.0028862222615215513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,2560,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,768,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,64,128,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,2048,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,512,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,256,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,1536,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,128,0.0018115554832749898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,1024,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,64,0.0018088887963030075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,768,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,225,32,32,0.0017928888814316855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,512,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,256,0.004861333303981357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,128,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,64,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,32,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,7168,0.010390222072601318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,225,32,5120,0.008635555704434713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,16384,0.40445955594380695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,16384,0.34245867199367946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,51200,1.2706044514973958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,12288,0.30871555540296763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,16384,0.1410799953672621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,12288,0.26366133160061306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,51200,0.4275520112779405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,10240,0.2684160073598226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,12288,0.10640177461836074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,10240,0.21366400188869902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,8192,0.21960888968573677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,10240,0.09037155575222439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,8192,0.17431110805935332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,51200,1.0998293558756511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,8192,0.07373599873648749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,7168,0.1891475518544515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,7168,0.1507937775717841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,10240,0.20234666930304634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,12288,0.2285217709011502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,7168,0.06556000312169392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,8192,0.15944000085194907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,6144,0.16292533609602186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,6144,0.05734399954477946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,5120,0.13792178365919325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,6144,0.13743555545806885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,7168,0.1419208844502767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,4096,0.1149839957555135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,5120,0.11044000254737006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,16384,0.31028445561726886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,5120,0.0484497778945499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,6144,0.11919466654459636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,3584,0.10071911414464314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,4096,0.09026666482289632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,4096,0.041152000427246094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,3072,0.08594577842288548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,3584,0.07939110861884223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,3584,0.03626399901178148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,2560,0.07225155830383301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,4096,0.08688088920381333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,3072,0.06898666752709283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,3584,0.0744311081038581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,3072,0.032829334338506065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,2048,0.05970933039983114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,3072,0.06674577792485555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,2560,0.059954669740464955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,2560,0.058033777607811816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,1536,0.04628088739183214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,2048,0.025045333637131587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,2048,0.05088000165091621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,1024,0.03288888931274414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,2048,0.050823999775780566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,1536,0.019754666421148512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,1536,0.03985155622164408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,1536,0.042972445487976074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,768,0.025822222232818604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,1024,0.015617777903874716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,1024,0.031298667192459106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,5120,0.10437955458958943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,512,0.018623999423450895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,768,0.01425866617096795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,768,0.025415110919210646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,768,0.029841777351167467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,256,0.010999110837777456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,512,0.021752888957659405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,512,0.025993777645958796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,256,0.018935110833909776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,2560,0.028538667493396338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,51200,0.9559155570136176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,128,0.009988444546858469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,128,0.01145244472556644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,256,0.021249777740902368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,64,0.008271111382378472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,128,0.0181697772608863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,65536,32,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,128,0.021324444148275588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,64,0.01793599956565433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,65536,32,0.017857778403494094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,65536,1024,0.03417511118782891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,512,0.01315822203954061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,65536,256,0.012085333466529846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,65536,1.1238088607788086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,51200,0.8594604598151313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,65536,1.0961475372314453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,16384,0.28007200029161244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,51200,0.839904891120063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,16384,0.27719200981987846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,65536,0.42333777745564777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,12288,0.21385333273145887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,16384,0.11177689499325222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,12288,0.2002248896492852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,10240,0.17674756050109863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,12288,0.08416444725460476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,51200,0.3355004522535536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,10240,0.07108977768156263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,8192,0.16700266467200386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,16384,0.25007200241088867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,10240,0.16740088992648652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,12288,0.19251022073957655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,7168,0.12449599636925592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,8192,0.138084438112047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,8192,0.05810844235950046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,10240,0.16005422009362116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,7168,0.12361778153313531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,6144,0.10871199766794841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,7168,0.051867554585138954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,6144,0.10372977786593968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,8192,0.13381511635250515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,6144,0.0454311105940077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,7168,0.11571644412146674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,5120,0.038711110750834145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,6144,0.10065244303809272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,4096,0.07706666655010648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,5120,0.08660799927181667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,3584,0.06613866488138835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,4096,0.07079022460513644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,4096,0.07237155569924249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,3584,0.06271377536985609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,5120,0.09222310781478882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,3072,0.05756533145904541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,3584,0.030057777961095173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,3584,0.06374310784869723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,2560,0.050865779320398964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,3072,0.055480887492497764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,5120,0.08733955356809829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,3072,0.057442665100097656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,51200,0.7495306862725152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,2560,0.024028443627887305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,2560,0.04802844590610928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,4096,0.0336151123046875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,2048,0.03998488850063748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,1536,0.03169066707293192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,2560,0.0507244434621599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,2048,0.02033244404527876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,1024,0.02387199964788225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,1536,0.03276000089115567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,2048,0.044109334548314415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,1536,0.03693422343995836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,768,0.0181404451529185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,1024,0.013843555417325763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,1024,0.02546666728125678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,1024,0.030663109487957422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,512,0.012460444536474017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,768,0.01274222218328052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,3072,0.026927999324268762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,768,0.02698488864633772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,256,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,512,0.011697777443461947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,512,0.017863111363516915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,512,0.022388445006476507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,128,0.007601777712504069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,256,0.01034755590889189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,256,0.016166221764352586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,256,0.019471999671724107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,2048,0.04278044568167793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,64,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,128,0.015150222513410779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,128,0.015211555692884656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,51200,65536,1.0304915110270183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,64,0.01516533394654592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,768,0.02072355482313368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,65536,0.38846755027770996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,65536,0.3552604516347249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,1536,0.01665866706106398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,51200,128,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,51200,0.3016604317559136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,65536,0.14526400301191542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,16384,0.10737866825527614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,51200,32,0.015036443869272867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,16384,0.09280888901816474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,51200,0.11270488633049859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,51200,0.281244436899821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,16384,0.040997332996792264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,12288,0.07726755407121447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,10240,0.06383822361628215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,16384,0.09046666489707099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,12288,0.032390223609076604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,12288,0.07087022066116333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,8192,0.05482844511667887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,12288,0.07095022333992852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,10240,0.027934223413467407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,10240,0.05931555562549167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,65536,0.3346115483178033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,7168,0.046712001164754234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,10240,0.061547557512919106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,51200,32,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,8192,0.04945422212282816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,8192,0.023778667052586872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,6144,0.04249155521392822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,8192,0.05221155616972181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,7168,0.044793777995639376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,5120,0.035756445593304105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,6144,0.018978666928079393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,51200,0.25956887669033474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,6144,0.038394666380352445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,7168,0.04643999867969089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,4096,0.029376887612872656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,6144,0.04212177793184916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,5120,0.016823111308945548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,3584,0.025772444076008264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,5120,0.0328382220533159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,5120,0.03746666510899862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,4096,0.015088000231319003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,4096,0.027906666199366253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,3072,0.023007111416922674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,4096,0.03250577714708116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,3584,0.024857777688238356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,2560,0.019139554765489366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,3072,0.013029333617952136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,3584,0.030060443613264296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,2048,0.015767110718621146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,3072,0.02731288969516754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,2560,0.01202400028705597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,2560,0.019291554888089497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,1536,0.012747555143303342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,2560,0.024549333585633173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,2048,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,2048,0.01695644524362352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,1024,0.010028444230556488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,7168,0.02141422198878394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,1536,0.008984888593355814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,1536,0.01443555619981554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,768,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,1024,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,1536,0.019479999939600628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,1024,0.011725333001878528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,512,0.006839111033413146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,1024,0.016689777374267578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,768,0.00739644467830658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,768,0.015185778339703878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,512,0.007231111327807109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,3584,0.01385244478782018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,256,0.004628444297446145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,3072,0.02204977803760105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,256,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,128,0.0038240001433425476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,512,0.012834666503800286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,256,0.00833511104186376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,64,0.003488000068399641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,256,0.011955555942323474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,16384,32,0.00350222239891688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,16384,128,0.006984000404675801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,128,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,128,0.009716444545321995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,16384,2048,0.022361778550677832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,64,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,32,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,65536,0.3033280107710096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,51200,0.2266906632317437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,65536,0.28987643453809947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,768,0.010344888601038191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,65536,0.12774754895104304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,16384,0.0760622223218282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,51200,0.22557245360480416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,16384,512,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,12288,0.05918044514126248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,51200,0.10155822171105279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,16384,0.03832533293300205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,16384,0.07673066854476929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,10240,0.05166844526926676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,12288,0.030046221282747056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,16384,0.08755911058849758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,12288,0.05910133653216892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,8192,0.04190399911668566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,12288,0.06856355402204725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,10240,0.02574399941497379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,10240,0.050247109598583646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,7168,0.037436445554097496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,10240,0.059296886126200356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,8192,0.021945777866575453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,8192,0.042025778028700084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,65536,0.31473154491848415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,6144,0.031785776217778526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,8192,0.050232887268066406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,7168,0.03726933399836222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,5120,0.027946667538748846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,6144,0.018005333013004728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,7168,0.04599377844068739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,6144,0.03279822071393331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,4096,0.02369244396686554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,6144,0.041189332803090416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,5120,0.02798933453030056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,5120,0.01613333324591319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,3584,0.020554666717847187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,5120,0.03609688745604621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,4096,0.014472888575659858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,4096,0.023387556274731953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,3072,0.01789688898457421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,4096,0.030934221214718286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,3584,0.013432888521088494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,3584,0.021309332715140447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,2560,0.015095111396577625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,3584,0.029140445921156142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,3072,0.012454222473833295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,51200,0.24839732382032606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,3072,0.018984888990720112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,2048,0.013058666553762225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,3072,0.026250667042202417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,2560,0.01146666705608368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,2560,0.0170817772547404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,1536,0.010695111420419482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,7168,0.019906666543748643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,2560,0.023775110642115276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,2048,0.010264889233642155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,1024,0.008930666579140557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,2048,0.014659555421935188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,1536,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,768,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,2048,0.02139733400609758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,1536,0.012727999852763282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,512,0.0058622223635514574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,1536,0.01902222302224901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,1024,0.010556444525718689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,768,0.007360888852013483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,256,0.0042053332759274375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,768,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,512,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,768,0.014150222142537435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,512,0.008207110895050896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,128,0.003568888952334722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,512,0.013091555900043912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,256,0.006824000014199152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,64,0.003432000055909157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,256,0.011355555719799466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,128,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,12288,32,0.0035013332962989807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,128,0.009370666411187913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,64,0.007277333074145847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,65536,0.2339155673980713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,32,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,51200,0.18736444579230416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,65536,0.24904976950751412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,12288,1024,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,65536,0.12441866927676731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,12288,1024,0.015443555182880826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,16384,0.06932089063856336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,51200,0.1921671099132962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,51200,0.09923289219538371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,256,0.007376000285148621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,12288,0.052405334181255765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,16384,0.0653466648525662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,16384,0.03649066554175483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,12288,128,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,10240,0.042280889219707914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,16384,0.08587466345893012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,12288,0.028841776980294123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,12288,0.05052533414628771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,8192,0.034191999170515276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,12288,0.0678773323694865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,10240,0.043199112017949425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,65536,0.31254667705959743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,10240,0.024501333634058636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,7168,0.031530665026770696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,51200,0.24382223023308647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,8192,0.021228444245126512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,8192,0.03632355398601956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,6144,0.02660888930161794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,7168,0.03211733367707994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,8192,0.050048000282711454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,7168,0.019262222780121695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,5120,0.022279111875428095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,7168,0.045088890526029796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,6144,0.01742755538887448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,6144,0.02794044547610813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,4096,0.01939911146958669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,5120,0.015832889411184523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,6144,0.04058311051792569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,5120,0.024088889360427856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,3584,0.016787555482652452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,4096,0.014442667365074158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,5120,0.03545777665244208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,4096,0.02049511174360911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,3072,0.014458666245142618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,4096,0.030701332622104224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,3584,0.018589332699775696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,2560,0.012753778033786349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,3584,0.028020444843504164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,3072,0.012154666913880242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,3072,0.017091555727852713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,2048,0.011041777829329172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,3072,0.025807999902301367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,2560,0.010693333215183683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,10240,0.05881333351135254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,1536,0.00923911150958803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,2560,0.02300888962215847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,2048,0.008901333643330468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,2048,0.013456000222100152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,1024,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,1536,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,2048,0.02107111116250356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,1536,0.011540444360838996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,768,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,1536,0.018597332967652213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,1024,0.009069333473841349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,1024,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,512,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,768,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,768,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,768,0.013551111022631327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,3584,0.013194666968451606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,256,0.0038373333712418876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,512,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,512,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,128,0.003173333282272021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,512,0.012477333347002665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,256,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,64,0.003148444410827425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,256,0.011403555671374003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,2560,0.01514933341079288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,128,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,10240,32,0.0031973332580592898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,128,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,64,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,32,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,65536,0.2039653327729967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,51200,0.15050133069356283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,65536,0.07749155494901869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,10240,1024,0.015428443749745687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,16384,0.05418489045566983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,51200,0.15802489386664495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,51200,0.06203999784257677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,10240,256,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,16384,0.054660442802641124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,12288,0.03979911075698005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,65536,0.17731644047631157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,10240,128,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,10240,0.03447022371821933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,12288,0.019314666589101154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,12288,0.0424177779091729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,65536,0.21314756075541177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,8192,0.028679112593332928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,12288,0.04253155655331082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,10240,0.01685422162214915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,10240,0.03628888726234436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,7168,0.025221332907676697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,8192,0.015107555521859063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,8192,0.030251555972629126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,8192,0.032460444503360324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,7168,0.026950221922662523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,6144,0.021536888347731695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,7168,0.02972266740269131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,6144,0.023800889650980633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,6144,0.012764444781674279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,5120,0.01865955524974399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,16384,0.023688889212078516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,6144,0.02684533264901903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,4096,0.015185778339703878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,5120,0.011749332977665795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,16384,0.052833778990639575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,51200,0.13901955551571316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,5120,0.020623111062579684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,3584,0.013450666434235044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,5120,0.023524444964196947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,4096,0.010948444406191507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,4096,0.021566222111384075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,3584,0.009679999616410997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,10240,0.03790488839149475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,3584,0.016283555163277518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,3584,0.02030755579471588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,2560,0.010421333213647207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,3072,0.009036444127559662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,7168,0.013791999883121915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,3072,0.015021334091822306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,2048,0.009296889106432596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,3072,0.019015999303923715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,2560,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,2560,0.013438222308953604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,1536,0.007861333588759104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,2560,0.017273777061038546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,2048,0.007595555649863348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,2048,0.01202400028705597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,1024,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,2048,0.016137777103318107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,1536,0.007083555890454187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,1536,0.010023111270533668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,768,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,1536,0.013923555612564087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,1024,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,1024,0.008255110846625434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,4096,0.01762488815519545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,512,0.0038328886859946777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,768,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,768,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,256,0.003257777748836411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,768,0.011065777805116443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,512,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,512,0.0069288888739215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,512,0.010840000377760993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,128,0.0028284444577164124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,256,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,256,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,64,0.0025395556456512874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,128,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,256,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,32,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,128,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,64,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,65536,0.17872977256774902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,8192,32,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,51200,0.14118133650885686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,65536,0.19691377215915254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,8192,1024,0.0122079998254776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,65536,0.07326666514078777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,8192,3072,0.012113778127564324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,16384,0.048244443204667836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,51200,0.05920088953442044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,51200,0.15420267317030165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,12288,0.040256887674331665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,16384,0.022838221655951604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,16384,0.05298666821585762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,10240,0.031889776388804116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,16384,0.051573332813051015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,8192,128,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,12288,0.018617777360810172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,12288,0.04120355513360765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,8192,0.027066666218969557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,51200,0.13346489270528158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,12288,0.041985776689317494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,7168,0.02431022293037838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,10240,0.016513778103722464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,10240,0.03512444429927402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,8192,0.014484445254007975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,10240,0.036526223023732506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,8192,0.02905511193805271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,6144,0.020594666401545208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,7168,0.01350400017367469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,8192,0.03166666626930237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,7168,0.026145777768558923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,7168,0.02884888980123732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,6144,0.023183999790085688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,4096,0.014867555763986377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,6144,0.025792888469166223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,5120,0.011391110718250275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,5120,0.020304888486862183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,3584,0.013208889298968844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,5120,0.02370311154259576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,4096,0.010028444230556488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,4096,0.017496888836224873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,3072,0.011754666765530905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,65536,0.16534754965040419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,3584,0.009019555317031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,3584,0.015879111157523263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,4096,0.021091555555661518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,2560,0.010696888797812991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,3584,0.01991555591424306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,3072,0.008370666868156856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,3072,0.014542222023010254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,2048,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,2560,0.008292444050312042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,3072,0.018333334061834548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,2560,0.013009777499569787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,1536,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,2560,0.017369776964187622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,2048,0.007377777662542131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,2048,0.011353777514563667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,5120,0.01754666699303521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,2048,0.015658666690190632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,1536,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,6144,0.012441777520709567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,1536,0.009415110780133141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,768,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,1536,0.013772444592581855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,1024,0.008046222229798635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,512,0.004518222063779831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,1024,0.011745777395036487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,768,0.006320000108745363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,768,0.011714666253990598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,256,0.003519111209445529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,512,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,512,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,512,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,128,0.003144888828198115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,256,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,256,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,256,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,64,0.0028035555862718155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,128,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,128,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,32,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,7168,128,0.00814844419558843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,7168,1024,0.006106666806671355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,64,0.006339555399285422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,65536,0.16436178154415554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,32,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,7168,1024,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,51200,0.13021332687801784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,65536,0.16014577282799616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,65536,0.07081955671310425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,16384,0.046606222788492836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,51200,0.12398310502370198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,51200,0.057146668434143066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,16384,0.043828444348441235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,12288,0.03710844450526767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,16384,0.02199111051029629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,65536,0.16206933392418757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,51200,0.12936000029246011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,10240,0.031066666046778362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,12288,0.01788444485929277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,12288,0.034359110726250544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,8192,0.025978666212823655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,12288,0.041160888142055936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,10240,0.016000888413853116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,10240,0.028951111767027114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,7168,0.022311111291249592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,10240,0.03625066743956672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,8192,0.014202665951516895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,8192,0.024572443630960252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,7168,0.013086222112178802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,7168,0.021987555755509272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,7168,768,0.0074648890230390764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,7168,0.028011557128694322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,5120,0.01662577854262458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,6144,0.012104888757069906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,6144,0.019698666201697457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,4096,0.014339556296666464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,6144,0.025621333056026038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,5120,0.011021333436171213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,5120,0.017254221770498488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,3584,0.01295644458797243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,4096,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,4096,0.015061333775520325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,16384,0.051065779394573636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,3072,0.011002666420406766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,4096,0.021009777983029682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,3584,0.009008888569143083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,3584,0.013824000126785703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,2560,0.010013333625263637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,3584,0.019256888164414298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,3072,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,3072,0.01275288893116845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,2048,0.008791111409664154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,2560,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,6144,0.01963644391960568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,2560,0.011671110987663269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,1536,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,8192,0.03080000148879157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,2560,0.01704088846842448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,2048,0.007328888608349695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,2048,0.010008888940016428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,1024,0.005535111245181825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,1536,0.006917333437336816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,2048,0.014880888991885714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,1536,0.008346666892369589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,768,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,1536,0.013429332938459186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,1024,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,1024,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,512,0.003526222374704149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,1024,0.011718221836619906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,5120,0.023016000787417095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,768,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,256,0.0029315555261241067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,768,0.01111288865407308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,512,0.00620888877246115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,512,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,512,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,256,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,256,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,256,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,3072,0.018188445104493033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,64,0.002502222235004107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,128,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,32,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,128,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,6144,128,0.008001777860853408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,65536,0.1460640033086141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,64,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,6144,32,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,51200,0.11593243810865615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,65536,0.15436444017622206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,65536,0.06792710887061225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,16384,0.043065776427586876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,6144,768,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,51200,0.05481422278616163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,65536,0.16011377175649008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,6144,128,0.0025315555847353404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,16384,0.0210506667693456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,12288,0.03263555632697211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,51200,0.12940088907877603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,16384,0.050438221957948476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,10240,0.028135110934575398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,12288,0.01742577718363868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,12288,0.03298755486806234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,8192,0.023352889551056757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,12288,0.040552000204722084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,10240,0.015506666567590503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,10240,0.028023110495673284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,10240,0.03549155592918396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,7168,0.019700444406933255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,8192,0.013794666363133324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,8192,0.023945776952637568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,6144,0.018381334013409086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,7168,0.02130577796035343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,8192,0.030512889226277668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,7168,0.012886222038004132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,51200,0.11977156003316243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,7168,0.027681777874628704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,6144,0.011701333026091257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,6144,0.019165333774354722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,4096,0.01401600076092614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,16384,0.04244977898067898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,5120,0.010371555884679159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,5120,0.01685599982738495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,3584,0.011902222202883827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,5120,0.023267555567953322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,4096,0.008941333327028487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,3072,0.010643555886215635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,4096,0.01480888823668162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,3584,0.008738666772842407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,4096,0.02033333314789666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,3584,0.013728000223636627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,2560,0.009673777553770278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,3584,0.019471999671724107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,3072,0.008059555457697975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,3072,0.012420444852775998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,2048,0.008061333662933773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,3072,0.01759466694460975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,2560,0.008038221961922115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,2560,0.0107004443804423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,1536,0.0068986668354935115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,2560,0.016347555650605094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,2048,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,1024,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,2048,0.01482399966981676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,1536,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,5120,0.015549333559142219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,1536,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,768,0.004226666771703296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,1536,0.01311466677321328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,1024,0.006501333167155583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,512,0.0035502223504914176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,1024,0.007280888656775157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,6144,0.025776000486479864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,768,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,1024,0.012120000190205045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,256,0.0031804444475306403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,768,0.006656000183688269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,512,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,768,0.011081777513027191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,128,0.002801777794957161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,256,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,512,0.010694222317801582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,256,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,64,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,256,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,5120,128,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,128,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,5120,128,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,65536,0.118724438879225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,64,0.0060159998635451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,32,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,51200,0.08530222045050727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,65536,0.11556800206502278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,51200,0.08922133180830215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,65536,0.15234577655792236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,16384,0.03496355480617947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,51200,0.12259644932217068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,512,0.0063582224150498705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,51200,0.05259644322925144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,12288,0.029487109846538965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,5120,32,0.00253866674999396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,16384,0.020270221763186984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,16384,0.048039999273088246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,12288,0.01683466633160909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,12288,0.025546666648652818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,10240,0.02437955637772878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,12288,0.03884711199336582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,8192,0.020232000284724765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,65536,0.0650577810075548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,10240,0.022379555635982092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,10240,0.03368355499373542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,7168,0.017809778451919556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,8192,0.013260444833172692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,8192,0.018921777606010437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,8192,0.029003553920321997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,6144,0.015440000428093804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,7168,0.01220533334546619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,7168,0.01718222267097897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,7168,0.026566222310066223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,5120,0.013444444371594323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,6144,0.010876444478829702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,6144,0.01568533314598931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,16384,0.03193333413865831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,4096,0.011422221859296164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,6144,0.02416977783044179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,5120,0.009932444327407414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,5120,0.013814222481515674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,3584,0.008982222113344405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,5120,0.021951999929216173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,4096,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,5120,2048,0.009335111412737105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,3072,0.008105778031879002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,3584,0.01847288840346866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,3072,0.010272888673676385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,2560,0.007131555842028723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,3072,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,3072,0.01703555550840166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,2560,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,2560,0.007625777688291337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,2048,0.006212444355090459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,2560,0.015627556376987033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,2048,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,2048,0.00810222244924969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,1536,0.005534222142563925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,4096,0.012039999994966718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,2048,0.01402311192618476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,1024,0.003939555750952827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,4096,0.01991555591424306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,3584,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,1536,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,1536,0.0127368892232577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,768,0.0036159998012913596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,1024,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,1024,0.011420444481902652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,512,0.0032222221295038858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,768,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,768,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,768,0.011394666300879585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,256,0.0028355556229750314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,512,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,512,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,512,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,128,0.002529777793420686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,256,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,256,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,256,0.009993777506881291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,64,0.002493333278430833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,128,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,128,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,4096,32,0.0025457777082920074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,10240,0.0151573336786694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,4096,128,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,4096,1536,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,65536,0.10455022255579631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,64,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,51200,0.08004355430603027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,32,0.005636444522274866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,1024,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,65536,0.06436089012357923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,16384,0.03231111168861389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,65536,0.11515200138092041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,51200,0.05122133427196079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,12288,0.027954667806625366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,51200,0.08854933579762776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,65536,0.1516853306028578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,16384,0.020045333438449435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,51200,0.12201600604587132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,16384,0.04678755667474535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,12288,0.016593777471118502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,12288,0.02516799999607934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,8192,0.019332443674405415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,10240,0.014839111102951897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,12288,0.037794666157828435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,10240,0.02184444500340356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,10240,0.032912890116373696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,7168,0.01690933273898231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,8192,0.012832889126406776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,8192,0.01887022289964888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,8192,0.028825776444541082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,6144,0.014766222900814481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,7168,0.011698666546079846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,7168,0.01715377800994449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,7168,0.0262044436401791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,5120,0.013040888640615674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,6144,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,6144,0.015527110960748462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,6144,0.02409422232045068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,16384,0.03154044349988302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,5120,0.009315555294354757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,5120,0.013787555197874704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,5120,0.021736888421906367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,3584,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,4096,0.008839999636014303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,4096,0.011701333026091257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,4096,0.019271110494931538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,3584,0.011193777951929303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,3072,0.00867377801073922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,4096,3584,0.011163555913501315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,3584,0.017878222796652053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,2560,0.007968000239796108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,3072,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,3072,0.016187555260128446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,2048,0.00685777763525645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,2560,0.009038222332795462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,2560,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,2560,0.015205333630243937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,1536,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,4096,0.010480889015727572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,2048,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,2048,0.008276444342401292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,10240,0.02318222158484989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,1024,0.004594666676388847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,2048,0.014144000079896716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,1536,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,768,0.004226666771703296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,1536,0.007333333293596904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,1024,0.006567111031876669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,1536,0.012442666623327466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,1024,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,512,0.0035075553589397003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,1024,0.011685333318180509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,768,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,768,0.006309333360857434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,512,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,512,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,512,0.010363555616802638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,3072,0.010095111197895473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,128,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,256,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,256,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,64,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,256,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,128,0.005776888794369168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,32,0.002247111043996281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,128,0.005730666634109285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,128,0.007335111498832703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,64,0.005752000129885144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3584,32,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,51200,0.06765244404474895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,65536,0.06312800116009183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,65536,0.10810222228368123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3584,3584,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,65536,0.15147466129726833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,51200,0.050272001160515674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3584,256,0.002551111082235972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,16384,0.028431998358832464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,51200,0.08518044153849284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3584,768,0.010507555471526252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,12288,0.0239955551094479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,16384,0.019780443774329293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,51200,0.12061777379777695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,10240,0.02163733376397027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,12288,0.016466667254765827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,16384,0.04651111033227709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,8192,0.01738666660255856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,12288,0.03754222061898973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,10240,0.014595554934607612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,10240,0.02145066691769494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,7168,0.01572355628013611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,10240,0.032497776879204646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,8192,0.01240355521440506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,8192,0.018400000201331247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,6144,0.014114666316244336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,8192,0.028170665105183918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,7168,0.011064888702498542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,7168,0.016713778177897137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,5120,0.011951111257076263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,7168,0.02618666655487484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,6144,0.010018666585286459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,6144,0.01516533394654592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,4096,0.009641778137948778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,6144,0.023954666323131982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,5120,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,5120,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,5120,0.021321778496106465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,3584,0.008046222229798635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,16384,0.030386666456858318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,4096,0.011448000040319232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,12288,0.024559110403060913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,4096,0.018522666560279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,3072,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,3584,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,3584,0.010371555884679159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,2560,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,65536,0.08209777540630765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,3584,0.017298666967286002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,3072,0.009640889035330879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,2048,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,3072,0.01614488826857673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,2560,0.007601777712504069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,2560,0.008712000317043727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,1536,0.004661333229806689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,2048,0.007270221908887227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,2560,0.015127110812399121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,2048,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,1024,0.0038213332494099936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,2048,0.01350044459104538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,1536,0.00675999994079272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,1536,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,768,0.003269333392381668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,1536,0.01242844429281023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,1024,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,4096,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,512,0.002875555513633622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,1024,0.011095110740926532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,768,0.006170666466156642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,768,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,512,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,512,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,3072,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,256,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,256,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,128,0.0023457778410779103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,256,0.009085333181752099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,128,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,64,0.0021679999513758556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,128,0.0053004444473319584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,128,0.0073520003093613525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,1024,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,65536,0.07246044609281752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,64,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,3072,32,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,3072,768,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,51200,0.05876533190409342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,65536,0.06183377901713053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,65536,0.10598933034472996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,256,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,16384,0.024067555864651997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,3072,512,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,65536,0.15084889199998644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,51200,0.04894222153557671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,51200,0.08354044622845119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,12288,0.020290666156344943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,16384,0.019705777366956074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,51200,0.12005244361029731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,10240,0.017829333742459614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,12288,0.015863110621770222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,16384,0.045964443021350436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,12288,0.02423911127779219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,8192,0.01404711107412974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,12288,0.03703555464744568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,10240,0.014108444253603617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,10240,0.021349334054523047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,3072,32,0.002285333350300789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,7168,0.012712000144852532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,10240,0.032123555739720665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,8192,0.012344000240166983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,6144,0.011040888726711273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,8192,0.018064889642927382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,7168,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,8192,0.02834133307139079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,5120,0.010018666585286459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,7168,0.016508445143699646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,6144,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,4096,0.010002666877375709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,6144,0.014654222461912366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,5120,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,6144,0.023550222317377727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,3584,0.009279111193286048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,5120,0.021191999316215515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,4096,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,4096,0.010727110836240979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,3072,0.008256889051861232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,4096,0.018600889378123812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,3584,0.008347555167145198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,3584,0.010199999643696679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,16384,0.0298942228158315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,3584,0.01719377769364251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,3072,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,3072,0.009389333426952362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,2048,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,3072,0.016146666473812528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,2560,0.008646222452322641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,2560,0.014470222923490735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,2048,0.00797155582242542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,2048,0.007186666958861881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,1536,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,2048,0.013417777915795645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,7168,0.025946666797002155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,1024,0.004206222378545337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,1536,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,5120,0.012420444852775998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,1536,0.007180444068378872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,768,0.003810666501522064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,1536,0.012732444538010491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,1024,0.006325333482689328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,2560,0.00720711135201984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,1024,0.01108888867828581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,768,0.006227555374304454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,512,0.003216888962520493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,768,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,256,0.002798222212327851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,768,0.011012444065676795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,512,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,2560,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,512,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,128,0.002209777840309673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,512,0.010281778044170803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,256,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,256,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,64,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2560,32,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,256,0.010058666268984476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2560,128,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,128,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,65536,0.057736886872185596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2560,128,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,51200,0.048004445102479726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,65536,0.06708266337712605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,65536,0.06066489219665527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,16384,0.020921778347757127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,1024,0.0064479998416370815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,65536,0.1500506665971544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,51200,0.04822044571240743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,12288,0.017824000782436795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,51200,0.11985777484046088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,16384,0.01924800044960446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,16384,0.020297777321603563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,10240,0.017117333081033494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,16384,0.04599110947714912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,12288,0.015459555718633862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,12288,0.01680088871055179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,8192,0.013284444808959961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,12288,0.03681866659058465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,10240,0.013074666261672974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,10240,0.015063110325071545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,7168,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,10240,0.032355556885401406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,8192,0.011367999845080905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,8192,0.013099555340078143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,6144,0.010703999963071613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,64,0.005535999933878581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,8192,0.027878221538331773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,7168,0.011701333026091257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,5120,0.009474666582213508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,7168,0.025548444853888616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,6144,0.010041777458455827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,51200,0.05262755685382419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,6144,0.023013333479563396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,4096,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,5120,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,5120,0.009304888546466827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,3584,0.007547555698288812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,4096,0.008627555436558193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,4096,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,4096,0.017887999614079792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,3584,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,3584,0.008252444366614023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,3584,0.01660977800687154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,3072,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,3072,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,2560,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,3072,0.015797333584891427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,7168,0.010727110836240979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,2048,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,2560,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,2560,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,6144,0.01036711119943195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,2560,0.014280888769361707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,1536,0.00424177779091729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,2048,0.007110222346252865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,2048,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,1024,0.0036924444139003754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,2048,0.013277333643701343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,1536,0.006308444258239534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,1536,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,5120,0.020255110330051847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,768,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,1536,0.012448000411192576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,1024,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,1024,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,3072,0.006685333533419504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,768,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,1024,0.011002666420406766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,768,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,768,0.010969777901967367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,256,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,512,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,512,0.005383111122581694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,128,0.0024906665914588505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,512,0.010363555616802638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,256,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2560,32,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,256,0.010006222460005019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,2048,128,0.005886222339338726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,32,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,128,0.005079999979999331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,2048,128,0.007589333587222629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,65536,0.05111022127999199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,64,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,32,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,51200,0.03903466794225905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,65536,0.05960533354017469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,512,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,65536,0.1503102249569363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,16384,0.017072889539930556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,51200,0.050795555114746094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,51200,0.046963555945290454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,51200,0.11965778138902451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,12288,0.014529777897728814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,16384,0.01964000033007728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,16384,0.01868355605337355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,2048,256,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,16384,0.0450524455971188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,12288,0.014793778459231058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,12288,0.016509332590632968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,12288,0.03643644518322415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,8192,0.013717333475748697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,10240,0.014487110906177096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,10240,0.012989333106411828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,10240,0.03177777926127116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,7168,0.012923555241690742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,8192,0.01143555591503779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,65536,0.06384978029463026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,8192,0.012424888710180918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,6144,0.011904000408119626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,8192,0.02736888825893402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,7168,0.010663111176755695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,7168,0.011419555379284753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,5120,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,6144,0.00998755544424057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,7168,0.02537511123551263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,6144,0.010334221853150262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,6144,0.02198133369286855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,5120,0.009277332988050249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,5120,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,3584,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,5120,0.01993688941001892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,4096,0.008617777791288164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,2048,64,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,4096,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,3072,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,4096,0.017648888958825003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,3584,0.008264000217119852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,2560,0.0064479998416370815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,3584,0.016588444511095684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,3072,0.007672888537247975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,3072,0.015398222539159985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,2048,0.005720888988839255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,2560,0.0076248885856734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,2560,0.00701422244310379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,2560,0.01442311041884952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,2048,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,1536,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,2048,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,2048,0.01311466677321328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,1536,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,1024,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,4096,0.007747555772463481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,1536,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,768,0.0035013332962989807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,1024,0.006322666588756774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,10240,0.013825777504179211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,1024,0.011081777513027191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,512,0.003173333282272021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,768,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,768,0.006132444573773279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,3584,0.008070222205585903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,256,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,3072,0.007340444458855524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,512,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,512,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,128,0.002163555473089218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,512,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,256,0.00573600000805325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,256,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,64,0.0019537777536445195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,256,0.009329777624871995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1536,128,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,128,0.00508088908261723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,65536,0.03649333450529311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,64,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,32,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,1536,0.01236977759334776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1536,1024,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,65536,0.06153421931796604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,65536,0.057546668582492404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,51200,0.029057777590221826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,768,0.011056000159846412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,65536,0.15035822656419542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,16384,0.014471999473041959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,51200,0.04866844415664673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,51200,0.1192435555987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,16384,0.016950221525298226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,12288,0.011807111402352652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,16384,0.01955911186006334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,16384,0.04506311151716444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1536,32,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,10240,0.011346666349305047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,12288,0.014093332820468478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1536,128,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,12288,0.015584889385435315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,8192,0.010647111468844943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,12288,0.036151111125946045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,10240,0.012777778009573618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,10240,0.013660444153679742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,7168,0.009881777895821465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,8192,0.011359111302428775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,8192,0.01184622198343277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,6144,0.010065777434243096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,8192,0.02642755541536543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,7168,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,7168,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,5120,0.009415110780133141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,7168,0.024383111132515803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,6144,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,6144,0.010120000276300643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,51200,0.04620888829231262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,4096,0.008247111406591203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,6144,0.02181333303451538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,5120,0.009272000028027428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,3584,0.007149332927332983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,5120,0.009318222602208456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,4096,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,5120,0.019892444213231403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,3072,0.006409777949253718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,4096,0.017266665895779926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,3584,0.008038221961922115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,3584,0.007927999728255803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,2560,0.005768888940413793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,3584,0.01641688909795549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,3072,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,2048,0.005199111170238919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,2560,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,3072,0.015499555402331881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,2560,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,1536,0.0042444442709287005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,10240,0.031304889255099826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,2048,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,2048,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,1024,0.0034799998005231223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,2048,0.013075555364290873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,1536,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,1536,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,768,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,1536,0.012441777520709567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,1024,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,1024,0.00565422202150027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,512,0.0027911110470692315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,768,0.005997333261701796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,1024,0.011029333704047732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,768,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,4096,0.008182221816645728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,256,0.0025191110455327565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,768,0.010729778144094678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,512,0.005886222339338726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,512,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,128,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,256,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,512,0.009925333162148794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,256,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,3072,0.007337777978844113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,256,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,1024,128,0.0057502223385704895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,32,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,128,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,65536,0.02974577744801839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,128,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,1024,2560,0.014317333698272705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,51200,0.023768888579474554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,64,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,1024,32,0.005071111023426056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,16384,0.014616888430383472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,65536,0.06020800272623698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,51200,0.04588177800178528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,51200,0.04740088846948412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,65536,0.14914222558339438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,12288,0.012050666742854647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,16384,0.01677688956260681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,51200,0.11860711044735378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,16384,0.018569777409235638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,10240,0.01131199962562985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,16384,0.04488355583614773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,12288,0.014096000128322177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,12288,0.03573866685231527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,8192,0.009745777481132084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,10240,0.012727999852763282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,10240,0.013471110827393003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,10240,0.0306640002462599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,1024,64,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,7168,0.009329777624871995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,8192,0.01108888867828581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,8192,0.011853333148691388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,6144,0.00849333322710461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,7168,0.010417777631017896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,7168,0.010983111129866706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,65536,0.0571004417207506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,7168,0.024319110645188227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,5120,0.007696888513035244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,6144,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,6144,0.009744889206356471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,4096,0.007352888584136963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,6144,0.021614222062958613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,5120,0.009257777697510189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,5120,0.008983111215962304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,3584,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,5120,0.019792889555295307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,4096,0.008610666626029545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,4096,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,3072,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,12288,0.014828445182906257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,4096,0.01738222274515364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,3584,0.008051555189821456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,2560,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,3584,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,3072,0.0076657773719893555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,3584,0.016505777835845947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,3072,0.0074675555030504865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,2048,0.004567111117972268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,3072,0.01538222200340695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,1536,0.004017777740955353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,2560,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,2048,0.007188444336255391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,8192,0.026084444589085046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,2048,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,1024,0.0034435554924938413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,2048,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,1536,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,1536,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,768,0.0031822222388452957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,1536,0.01203644441233741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,1024,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,1024,0.005648889061477449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,512,0.002838222309947014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,1024,0.011031111081441244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,768,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,256,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,768,0.010695111420419482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,512,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,512,0.0052968888647026485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,128,0.0022071111533376905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,256,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,256,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,256,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,128,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,64,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,768,32,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,128,0.005048888838953442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,65536,0.024753777517212763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,768,2560,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,128,0.007380444142553542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,2560,0.014463999205165438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,32,0.005116444494989184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,51200,0.01832177738348643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,65536,0.05678844451904297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,65536,0.05839733282725016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,16384,0.010688888529936472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,51200,0.04474311073621114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,65536,0.1486862235599094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,51200,0.046945777204301625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,12288,0.012648888760142855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,768,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,16384,0.016544888416926067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,51200,0.1189146704143948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,16384,0.018048889107174344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,16384,0.04361777835422092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,768,512,0.009683555199040307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,12288,0.014647111296653748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,8192,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,12288,0.035143110487196184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,10240,0.01311555587583118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,10240,0.012757333616415659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,7168,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,10240,0.030206223328908283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,8192,0.011612444288200803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,8192,0.026032888227038916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,768,64,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,7168,0.01071911139620675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,6144,0.008736888567606607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,7168,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,5120,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,7168,0.023805333508385554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,6144,0.009710222482681274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,6144,0.00996800015370051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,6144,0.021911111142900255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,4096,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,5120,0.009297777381208207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,10240,0.011343110766675739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,12288,0.014096889230940076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,3584,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,4096,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,4096,0.017498667041460674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,3584,0.008265777594513362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,3584,0.007671111159854465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,3072,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,3584,0.016488000750541687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,3072,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,3072,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,2560,0.004987555659479565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,3072,0.015441776977645027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,2560,0.0073724447025193115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,2560,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,2048,0.004629333400064045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,5120,0.009033777647548253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,5120,0.01943644384543101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,1536,0.004181333300140169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,2048,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,2048,0.006308444258239534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,4096,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,2048,0.013399110900031196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,1024,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,1536,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,1536,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,768,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,8192,0.011063999599880643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,1536,0.012104888757069906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,1024,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,512,0.0027022223091787765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,1024,0.011041777829329172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,768,0.00612088872326745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,768,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,256,0.0024115554988384247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,768,0.010399999717871347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,512,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,128,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,512,0.009685333404276106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,256,0.00563733321097162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,256,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,64,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,256,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,2560,0.014461333552996317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,512,32,0.002151111140847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,128,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,128,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,65536,0.018167111608717177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,512,128,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,64,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,51200,0.015430221954981485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,65536,0.05079911152521769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,32,0.005048000150256687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,65536,0.05729777945412529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,16384,0.00998755544424057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,51200,0.039750221702787615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,65536,0.14869599872165257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,51200,0.046118223004870944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,12288,0.008458666503429413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,512,1024,0.005886222339338726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,51200,0.11750400066375732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,16384,0.016568000117937725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,16384,0.01777688827779558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,10240,0.007780444290902879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,16384,0.04384444488419426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,12288,0.014074666632546319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,8192,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,12288,0.01461955573823717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,10240,0.012662221988042196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,512,512,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,7168,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,12288,0.03499199946721395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,10240,0.013063111239009433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,6144,0.006588444527652528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,10240,0.029873778422673542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,8192,0.01108088923825158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,8192,0.011413333316644033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,7168,0.01052266690466139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,5120,0.006897777732875612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,8192,0.026199999782774184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,6144,0.009678222239017487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,4096,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,7168,0.023395554886923894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,6144,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,5120,0.0092613332801395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,6144,0.021706667211320665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,3584,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,5120,0.009020444419648912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,4096,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,5120,0.01924888955222236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,3072,0.005521777603361342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,4096,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,3584,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,4096,0.01722577710946401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,3584,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,3072,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,3584,0.016532444291644625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,3072,0.00740622232357661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,2048,0.0041928887367248535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,3072,0.01478666729397244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,2560,0.007557333343558841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,2560,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,1536,0.0038773334688610504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,2560,0.01407199932469262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,2048,0.007055111229419708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,2048,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,1024,0.003154666680428717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,2048,0.013071111506885953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,1536,0.00665155549844106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,1536,0.006103111224042044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,768,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,7168,0.010728889041476779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,1536,0.011766222616036734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,1024,0.006302222195598815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,1024,0.005773333211739858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,512,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,1024,0.010685332947307162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,768,0.005379555539952383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,256,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,768,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,512,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,2560,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,128,0.0021013333979580137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,256,0.005858666780922148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,256,0.00508266646001074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,256,0.00925422211488088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,64,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,128,0.00563733321097162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,128,0.005022222383154763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,128,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,65536,0.014939554863505893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,64,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,256,32,0.004953777624501122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,51200,0.013095999757448832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,65536,0.049154665735032826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,65536,0.05617244376076592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,768,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,65536,0.14754399988386366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,51200,0.03978577918476529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,256,512,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,256,512,0.010209778116808997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,16384,0.017551110850440133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,16384,0.016523554921150208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,16384,0.04353511002328661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,256,32,0.0018684443914228016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,12288,0.014111111561457316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,10240,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,12288,0.03456888927353753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,10240,0.012616888516479068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,10240,0.01309333327743742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,8192,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,10240,0.029828445778952703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,8192,0.01110311100880305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,8192,0.011409777734014722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,7168,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,51200,0.044285333818859525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,8192,0.025690666503376428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,6144,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,7168,0.010592000352011787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,12288,0.00665244460105896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,7168,0.010679110884666443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,5120,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,7168,0.02344977855682373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,6144,0.009669333696365356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,6144,0.009679111341635386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,4096,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,6144,0.021987555755509272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,12288,0.014497778481907315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,5120,0.008990222381220924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,5120,0.008747555315494537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,3584,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,5120,0.018979556030697294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,4096,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,3072,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,3584,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,3584,0.007621333003044128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,3584,0.015850666496488783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,2560,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,3072,0.007881777981917063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,3072,0.0070053330726093715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,16384,0.007244444555706448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,3072,0.015152000718646579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,2560,0.007326222128338284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,2560,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,51200,0.11725600560506184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,1536,0.0035128887328836652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,2560,0.014246222045686511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,2048,0.0063075555695427795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,2048,0.007037333316273159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,1024,0.003170666595300039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,2048,0.01275644451379776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,1536,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,1536,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,768,0.002797333316670524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,1536,0.012030222349696688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,1024,0.006312888943486744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,1024,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,1024,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,4096,0.008030222521887885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,768,0.005970666805903117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,512,0.002493333278430833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,768,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,4096,0.01757333344883389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,256,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,768,0.010370666782061258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,512,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,128,0.0019404445257451797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,512,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,256,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,512,0.009668444593747457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,256,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,256,0.009655111365848118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,224,128,128,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,65536,0.01124888906876246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,64,0.004955555415815777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,51200,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,32,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,16384,0.005843555761708154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,65536,0.054839111036724515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,12288,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,16384,0.017511111166742113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,10240,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,12288,0.01456622282663981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,8192,0.005694222119119432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,10240,0.013056889176368713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,7168,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,64,0.0018044445249769424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,6144,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,224,128,128,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,128,128,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,8192,0.011456888582971362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,5120,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,4096,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,3584,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,7168,0.010666666759385003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,3072,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,6144,0.009537777966923183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,2560,0.004874666531880696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,5120,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,4096,0.007817777494589487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,2048,0.004952888935804367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,1536,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,3584,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,1024,0.002838222309947014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,3072,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,768,0.0027919999427265594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,51200,0.043939554029040866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,512,0.002500444443689452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,1536,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,256,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,1024,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,128,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,512,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,64,32,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,256,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,65536,0.011244444383515252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,128,0.00499288903342353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,64,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,51200,0.00960177762640847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,128,2048,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,16384,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,12288,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,32,0.004951111144489712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,10240,0.005660444498062134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,65536,0.05437066819932726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,51200,0.043826666143205434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,8192,0.005655111124118169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,16384,0.017498667041460674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,7168,0.00535111129283905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,6144,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,12288,0.014730667074521383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,5120,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,10240,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,4096,0.004577777865860197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,8192,0.011325333681371478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,3584,0.005092444519201915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,7168,0.010382222632567087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,2560,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,6144,0.009582222335868413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,3072,0.004858666823969947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,2560,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,5120,0.008984888593355814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,2048,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,1536,0.0035146665241983202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,4096,0.008040000167157915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,1024,0.0028568889117903183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,768,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,3584,0.007431999676757389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,768,0.0026906666656335196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,3072,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,512,0.0024702221983008915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,2560,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,256,0.0021226666867733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,2048,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,1536,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,1024,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,64,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,512,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,256,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,128,0.004770666774776247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,64,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,32,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,64,2048,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,128,0.0018604445374674266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,16384,0.3930533462100559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,224,32,768,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,224,32,32,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,51200,1.2445742289225261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,51200,1.0485306845770943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,12288,0.30071110195583767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,16384,0.3345457712809245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,16384,0.13857244120703802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,10240,0.25006045235527885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,51200,0.4133591122097439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,8192,0.2124720017115275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,10240,0.08970044718848334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,10240,0.2111297713385688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,8192,0.17014666398366293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,8192,0.07198844353357951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,7168,0.1767564482159085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,6144,0.1542159981197781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,7168,0.06341155370076497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,7168,0.15131733152601454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,5120,0.12829332881503636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,6144,0.13172445032331678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,6144,0.0557733310593499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,5120,0.047057777643203735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,5120,0.10832355419794719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,3584,0.09195378091600205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,4096,0.08981066942214966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,12288,0.24979376792907715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,4096,0.040592888991038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,3072,0.08003111018074884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,3584,0.03577422102292379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,2560,0.06697155369652642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,3584,0.07862844732072619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,2048,0.054494222005208336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,3072,0.0321324434545305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,3072,0.06815466615888807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,2560,0.028054220808876887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,2560,0.059338668982187905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,12288,0.10497511095470852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,2048,0.024420445164044697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,2048,0.049878223074807056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,1536,0.019104000594880845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,768,0.02404444416364034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,1536,0.03933866818745931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,1024,0.015856888559129503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,512,0.01739644507567088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,1024,0.030831111801995173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,256,0.011591999895042844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,4096,0.10661333137088352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,512,0.013432888521088494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,512,0.025736888249715168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,128,0.009119110802809397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,768,0.024977778395016987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,256,0.01891377733813392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,64,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,256,0.011738667057620155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,32,0.00757422215408749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,128,0.01166044423977534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,128,0.018184888694021437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,64,0.017883555756674874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,65536,32,0.018020444446139865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,1536,0.043952888912624784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,65536,1024,0.03127377894189622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,65536,768,0.014126222994592456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,16384,0.322160005569458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,65536,1.0944684346516926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,51200,0.9968080520629883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,65536,1.288871129353841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,51200,0.8169377644856771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,16384,0.11008533504274155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,51200,0.32619555791219074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,10240,0.20068266656663683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,12288,0.24435199631585014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,65536,0.4171146551767985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,8192,0.16421600182851157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,12288,0.2057537767622206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,10240,0.0697173343764411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,8192,0.05759021970960829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,10240,0.17131022612253824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,7168,0.1430551078584459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,6144,0.12439110543992783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,16384,0.26679910553826225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,8192,0.13663200537363687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,7168,0.11874755223592122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,5120,0.10454577869839138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,6144,0.10275110933515762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,4096,0.08576355377833049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,5120,0.03836799992455377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,3584,0.0751368867026435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,4096,0.03301066822475857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,5120,0.08644266923268636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,4096,0.07135288582907783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,6144,0.0454328887992435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,3072,0.0657715532514784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,3584,0.029531555043326482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,3584,0.06314400169584486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,2560,0.055028445190853544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,2048,0.044813334941864014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,3072,0.02651111119323307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,3072,0.05471377902560764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,2560,0.023920888702074688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,12288,0.08190488815307617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,1536,0.035014222065607704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,2560,0.04785155587726169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,1024,0.025448888540267944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,1536,0.01645155582163069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,1536,0.032376001278559365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,768,0.019742222295867074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,7168,0.05091911223199633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,1024,0.025573333104451496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,512,0.014119999276267158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,768,0.012864000267452665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,256,0.009667555491129557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,512,0.011735999749766456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,768,0.020280000236299302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,128,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,512,0.02100177771515316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,256,0.010392000277837118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,64,0.006165333506133821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,256,0.015916445189052157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,128,0.010371555884679159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,51200,32,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,128,0.015108444624476962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,64,0.0151235560576121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,32,0.015152888165579902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,65536,0.3790408770243327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,2048,0.020136000381575692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,51200,2048,0.040588445133633085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,51200,1024,0.013806222213639153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,51200,0.30395110448201496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,65536,0.1419315603044298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,65536,0.3544986777835422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,16384,0.09081955750783284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,16384,0.04014399978849623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,51200,0.2792195479075114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,51200,0.11145955986446804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,10240,0.06242399745517307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,8192,0.05278844303554959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,12288,0.03194844391610887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,10240,0.02773688899146186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,12288,0.06874133480919732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,7168,0.04554133282767402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,10240,0.0581715570555793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,8192,0.02379111117786831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,6144,0.03911022345225016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,8192,0.04913155568970574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,5120,0.03391022152370877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,7168,0.02130311065249973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,7168,0.04303555687268575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,6144,0.019187556372748483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,6144,0.0376782218615214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,5120,0.016843555702103507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,5120,0.03258400162061056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,4096,0.015146666102939181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,3072,0.022452443838119507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,4096,0.027777777777777776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,3584,0.013796444568369122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,16384,0.09558844566345215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,2560,0.018453333112928603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,3584,0.02477688921822442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,3072,0.012788444757461548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,12288,0.07479200098249647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,3072,0.021967111362351313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,1536,0.012157333393891653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,2560,0.01204622205760744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,2560,0.01921866668595208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,2048,0.010993777877754636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,1536,0.008966222405433655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,1024,0.009382222261693742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,768,0.007883555359310573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,1536,0.014182221558358936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,1024,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,512,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,1024,0.011855111353927188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,768,0.007631111476156447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,256,0.004263111286693149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,768,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,4096,0.028034667174021404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,128,0.003608889049953885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,3584,0.025037333369255066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,256,0.00848711116446389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,64,0.0032097777972618737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,128,0.006836444553401735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,32,0.0035084444615576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,16384,2048,0.01514488955338796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,64,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,32,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,2048,0.016926222377353244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,65536,0.29067643483479816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,51200,0.2269137700398763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,65536,0.2860151131947835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,65536,0.1268764469358656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,512,0.007255111303594377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,16384,0.08191021945741442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,16384,256,0.006743999818960826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,12288,0.05858933263354831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,51200,0.10114222102695042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,128,0.008178667061858708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,51200,0.22426311175028482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,16384,0.07607911030451457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,10240,0.05185688866509331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,8192,0.042540444268120654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,12288,0.03006311257680257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,10240,0.02567200031545427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,12288,0.057879110177357994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,10240,0.0494959983560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,8192,0.02179199953873952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,8192,0.04162844353251987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,6144,0.03231199913554721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,7168,0.019913777709007263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,5120,0.027316444449954565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,6144,0.01806400054030948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,6144,0.03254222207599216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,4096,0.023281777898470562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,5120,0.016102222932709586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,5120,0.02779910961786906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,3584,0.020263110597928364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,16384,512,0.0098213329911232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,4096,0.014537778165605334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,3072,0.018027555611398485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,4096,0.02336799932850732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,3584,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,16384,0.03779644436306424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,2560,0.015136889285511441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,3584,0.021049777666727703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,2048,0.012934221989578672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,3072,0.012712889247470431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,1536,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,3072,0.019107555349667866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,2560,0.017039111918873258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,7168,0.037738667594061956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,1024,0.008640888664457533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,1536,0.008355555435021719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,768,0.007257777783605788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,1536,0.012652444342772165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,1024,0.007763555480374231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,512,0.005701333284378052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,1024,0.010053333308961656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,768,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,256,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,768,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,512,0.0070293330483966405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,128,0.003499555504984326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,512,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,256,0.006656000183688269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,256,0.007332444190979004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,128,0.006663110935025745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,32,0.0032186667538351486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,128,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,64,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,2560,0.011403555671374003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,32,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,65536,0.2680693202548557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,2048,0.014751111467679342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,51200,0.20532711346944174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,16384,0.07598933246400622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,12288,7168,0.03648533423741659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,65536,0.12410310904184978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,51200,0.1927271154191759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,51200,0.09828533066643609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,12288,0.059953775670793324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,10240,0.049215111467573375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,16384,0.03645155496067471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,12288,0.028578665521409776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,16384,0.06525599956512451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,12288,64,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,12288,0.050028443336486816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,8192,0.04136177897453308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,10240,0.024901333782407973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,7168,0.03864799936612447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,10240,0.042600890000661217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,6144,0.03237688872549269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,8192,0.021220443977249995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,7168,0.019375110665957134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,8192,0.03604088889227973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,6144,0.017471111483044095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,7168,0.031593776411480375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,4096,0.022416889667510986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,6144,0.027748444014125403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,12288,2048,0.009739555418491364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,3584,0.020149333609475028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,5120,0.01583377851380242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,3072,0.017247110605239868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,4096,0.01406133340464698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,5120,0.024030221833123103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,4096,0.020554666717847187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,2560,0.014757333530320061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,3584,0.018572444717089336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,3072,0.012078222301271228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,65536,0.2426053418053521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,2048,0.012652444342772165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,2560,0.010597333312034607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,3072,0.016873778568373788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,1536,0.010468444062603844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,2560,0.015224888920783997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,2048,0.008799110849698385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,1024,0.008143111235565608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,2048,0.013251555462678274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,1536,0.008119111259778341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,768,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,1536,0.011400000088744693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,512,0.00508088908261723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,1024,0.009091555244392818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,768,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,256,0.003920000046491623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,768,0.008303110798199972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,512,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,5120,0.026632888449562922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,128,0.0032479998966058097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,512,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,256,0.006593777901596493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,128,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,32,0.00350577798154619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,128,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,3584,0.013057777451144325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,64,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,32,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,65536,0.22088266743554008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,51200,0.17210400104522705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,65536,0.07670222388373481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,65536,0.19729777177174887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,10240,1024,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,16384,0.061965333090888135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,12288,0.04556533363130358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,51200,0.15450133217705622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,16384,0.02385777731736501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,10240,0.038725333081351385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,51200,0.06066044171651205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,16384,0.054085334142049156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,12288,0.019280888968043856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,10240,64,0.0031866667171319327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,12288,0.04145866632461548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,10240,0.017024000485738117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,10240,256,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,7168,0.02771200074089898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,10240,0.035975999302334256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,6144,0.024420445164044697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,8192,0.029543999168607924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,5120,0.02089599933889177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,6144,0.012792888614866467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,7168,0.026371555195914373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,4096,0.017820444371965196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,6144,0.023183999790085688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,5120,0.011707555916574268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,3584,0.015047111445003085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,5120,0.02034311162100898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,3072,0.013393777940008374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,3584,0.009849777652157677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,4096,0.01756977703836229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,2560,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,3072,0.008842666943868002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,3072,0.01481599940194024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,2048,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,2560,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,2560,0.013420444395807056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,1536,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,8192,0.032079110542933144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,2048,0.007744889292452071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,2048,0.011852444046073489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,1024,0.006551110910044775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,8192,0.014904000692897372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,1536,0.007138667007287343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,768,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,1536,0.009787555370065901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,7168,0.013797333670987023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,512,0.004137777619891697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,1024,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,256,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,1024,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,768,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,128,0.0028568889117903183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,512,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,768,0.007592889169851939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,64,0.002521777732504739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,256,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,512,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,8192,32,0.0028239999794297745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,4096,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,256,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,128,0.006333333336644703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,3584,0.016356445021099515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,64,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,65536,0.20525422361161974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,8192,32,0.006637333167923822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,51200,0.16288267241583929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,65536,0.07213866710662842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,65536,0.19453155994415283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,12288,0.04262666569815742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,51200,0.058780444992913135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,16384,0.02264533274703556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,51200,0.1518933375676473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,16384,0.05271288752555847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,10240,0.03616177704599168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,12288,0.018516444497638278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,8192,0.030279109875361126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,12288,0.04044444362322489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,7168,0.026531555586391028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,10240,0.03475111060672336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,8192,0.029151999288135107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,7168,0.013627555635240344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,7168,0.025731555289692346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,8192,128,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,5120,0.01937955617904663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,6144,0.012415111064910889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,6144,0.022794667217466567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,4096,0.01605422298113505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,5120,0.011428444749779172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,5120,0.019948444432682462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,3584,0.014296000202496847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,4096,0.010056888891590966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,4096,0.01722666621208191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,3072,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,3584,0.009280888570679558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,3584,0.015728889240158927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,2560,0.011051555474599203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,3072,0.00867911097076204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,8192,0.014599111345079211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,3072,0.014366222752465142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,2048,0.009862222605281407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,2560,0.007920000288221572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,6144,0.022794667217466567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,1536,0.00794488853878445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,2048,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,1024,0.00619555554456181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,2048,0.011074666347768573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,1536,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,768,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,1536,0.009315555294354757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,1024,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,512,0.004179555508825514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,1024,0.008082666330867344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,16384,0.054824001259273954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,768,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,256,0.003346666693687439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,768,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,128,0.0029102222373088202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,512,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,512,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,64,0.002700444517864121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,7168,32,0.0028817777832349143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,256,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,128,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,256,0.006350222147173352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,128,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,7168,10240,0.016400888562202454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,65536,0.17279288503858778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,64,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,32,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,16384,0.05156444509824117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,51200,0.1417484415902032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,65536,0.0698088871108161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,65536,0.15584622489081487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,12288,0.04177688890033298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,7168,2560,0.013066666821638743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,10240,0.03525422347916497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,51200,0.12171199586656357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,51200,0.05641333262125651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,16384,0.04316622349951002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,8192,0.02964977754486932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,12288,0.018009778526094224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,7168,0.02588622272014618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,12288,0.03339644604259067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,10240,0.01588622232278188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,6144,0.02420177724626329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,8192,0.014164444473054675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,5120,0.018956444329685636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,8192,0.023856000767813787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,7168,0.013082666529549493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,4096,0.015388444066047668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,7168,0.021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,5120,0.010775999890433418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,6144,0.019369777705934312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,3584,0.013607111242082385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,4096,0.009460444251696268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,5120,0.017065778374671936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,3072,0.012065778175989786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,4096,0.014979556202888489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,3584,0.009015999734401703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,2560,0.01053600013256073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,3584,0.013616888887352414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,3072,0.008119111259778341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,2048,0.008792888787057664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,3072,0.012746666868527731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,2560,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,1536,0.007459555235173967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,2560,0.01128799964984258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,2048,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,1024,0.005757333503829108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,2048,0.009810667071077559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,1536,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,16384,0.021592888567182753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,768,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,1024,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,1536,0.00832799987660514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,512,0.0038453332251972626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,1024,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,256,0.0031786666562159858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,512,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,768,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,10240,0.02830933199988471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,128,0.002711999954448806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,512,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,64,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,256,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,6144,32,0.0025315555847353404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,128,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,128,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,64,0.005974222388532426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,65536,0.1336159971025255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,32,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,51200,0.11144799656338161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,65536,0.0673475530412462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,65536,0.15052266915639242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,16384,0.04916622241338094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,51200,0.11797066529591878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,51200,0.054409778780407377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,12288,0.03535288903448317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,16384,0.02105866703722212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,16384,0.042019555966059365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,768,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,12288,0.03218844532966614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,8192,0.02754488918516371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,10240,0.015531554818153381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,6144,6144,0.012136889000733694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,7168,0.024217777782016333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,10240,0.0277502222193612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,6144,256,0.005992888990375731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,6144,0.023423110445340473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,8192,0.013783999615245394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,8192,0.02366488840844896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,7168,0.012730666332774691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,5120,0.018810666269726224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,6144,0.011711999773979187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,7168,0.021033777130974665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,4096,0.015915556086434256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,6144,0.018766222728623282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,5120,0.0103493332862854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,3584,0.0144213338692983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,4096,0.00903911143541336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,5120,0.01681511104106903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,3072,0.010764444039927589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,4096,0.01457333399189843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,2560,0.009327111144860586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,3584,0.013732444081041547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,3072,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,2048,0.008176888856622908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,3072,0.012392889294359418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,10240,0.033238222201665245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,2560,0.007557333343558841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,1536,0.006888888776302338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,12288,0.01738577749994066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,1024,0.00499911109606425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,2560,0.010762666662534079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,2048,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,768,0.004222222086456087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,1536,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,2048,0.009343110852771336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,512,0.0035626664757728577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,1536,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,768,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,1024,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,512,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,768,0.006670222100284364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,128,0.002833777831660377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,256,0.005705777969625261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,512,0.006404444575309753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,64,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,256,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,32,0.002817777709828483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,128,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,128,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,65536,0.11818222204844157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,64,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,5120,32,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,51200,0.09501689010196263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,65536,0.06492355797025892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,65536,0.11141867107815212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,51200,0.052108443445629544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,51200,0.0886933339966668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,12288,0.029080000188615587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,16384,0.0314026673634847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,10240,0.02440533373090956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,12288,0.016837333639462788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,1024,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,12288,0.025256888733969793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,5120,256,0.0029075555503368378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,10240,0.015154666370815702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,7168,0.017674666312005784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,10240,0.02179644505182902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,8192,0.013232889274756113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,8192,0.018574222922325134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,6144,0.015535111228624979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,7168,0.016935999194780987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,5120,3584,0.008359111017651027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,6144,0.010988444089889526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,5120,0.013561777770519257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,6144,0.015309333801269531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,16384,0.037055111593670316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,4096,0.015827554795477126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,5120,0.009678222239017487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,3584,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,5120,0.013607111242082385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,16384,0.020216888851589628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,3072,0.011036444041464063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,3584,0.008290666672918532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,8192,0.020234665936893888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,2560,0.009958222508430481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,3072,0.010123555858929953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,2560,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,2560,0.008864888714419471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,2048,0.008546666966544257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,7168,0.012057777908113269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,1536,0.0070017774899800616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,2048,0.007255111303594377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,2048,0.008007110820876228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,1024,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,1536,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,1536,0.007324444750944774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,768,0.004559111264016894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,1024,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,4096,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,512,0.004047111090686587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,768,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,768,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,4096,0.012147555748621622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,256,0.0028293333533737394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,3584,0.011375111010339526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,512,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,128,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,3072,0.00793511089351442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,256,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,64,0.0024560000747442245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,4096,32,0.002471111093958219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,4096,128,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,128,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,65536,0.09222310781478882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,64,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,32,0.005879111174080107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,51200,0.08266755607393053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,16384,0.036010666026009455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,65536,0.06400089131461249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,65536,0.11046399672826131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,51200,0.0509422222773234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,51200,0.08711022138595581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,16384,0.019946666227446664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,1024,0.006568888823191325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,12288,0.01677688956260681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,12288,0.024728889266649883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,8192,0.01900888813866509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,10240,0.014906666345066495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,512,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,10240,0.021537777450349595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,7168,0.016885333591037326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,8192,0.012940444052219391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,8192,0.018566222654448617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,6144,0.015040000279744467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,7168,0.011752000285519494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,5120,0.013097777962684631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,7168,0.016875555117925007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,6144,0.010412444670995077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,4096,0.011729777687125735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,6144,0.015226667126019796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,5120,0.009306666751702627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,3584,0.00981155534585317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,5120,0.013788444300492605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,4096,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,3072,0.008725333544943068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,4096,0.011753777662913004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,3584,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,2560,0.007893333004580604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,3584,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,3072,0.008039111064540016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,10240,0.023058666123284235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,2048,0.006743999818960826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,3072,0.009694221946928237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,2560,0.007436444362004598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,2560,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,2048,0.007349333001507654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,1024,0.004628444297446145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,2048,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,1536,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,1536,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,768,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,4096,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,1024,0.006351111249791251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,1024,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,768,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,256,0.0026853332916895547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,768,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,512,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,512,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,128,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,256,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,256,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,64,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,12288,0.028601778878106013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3584,128,0.005807111246718301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,128,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,16384,0.03125155634350247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,65536,0.08226755592558119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,64,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,1536,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3584,32,0.00563733321097162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,51200,0.06938933663898043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,16384,0.027015111512607996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,65536,0.06263289186689588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,51200,0.04991111159324646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,51200,0.08334044615427653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,512,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,12288,0.02258666687541538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,16384,0.01978755493958791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,10240,0.022209778428077698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,16384,0.030284444491068523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,12288,0.01633777717749278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,8192,0.01735111077626546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,12288,0.024039111203617517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,10240,0.014518222875065274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,7168,0.01593955523437924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,10240,0.020961778031455144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,6144,0.015251555376582675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,8192,0.018062222335073683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,7168,0.011396444506115384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,5120,0.012626666989591388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,7168,0.01663555536005232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3584,32,0.0022088889446523455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,6144,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,4096,0.009710222482681274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,6144,0.014936889211336771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,3584,0.009923555784755284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,5120,0.009381333159075843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,5120,0.013080000049538083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,4096,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,3072,0.008615111311276754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,4096,0.011062222222487131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,2560,0.0075431110130416016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,3584,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,65536,0.10645066367255317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,3072,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,2048,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,3072,0.009560888840092553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,1536,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,2560,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,2560,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,1024,0.00408000002304713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,2048,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,1536,0.006650666809744305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,768,0.0035208890007601846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,1536,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,1024,0.0064071110553211635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,8192,0.012250666817029318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,512,0.0030177777840031516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,256,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,768,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,512,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,128,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,512,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,256,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,64,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,256,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,3072,32,0.0021520000365045336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,128,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,128,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,3584,0.008251555263996124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,65536,0.07950755622651842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,64,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,32,0.005807111246718301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,65536,0.06136711438496908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,65536,0.10410488976372613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,16384,0.025638222694396973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,3072,2048,0.007340444458855524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,51200,0.049088888698154025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,12288,0.020655110478401184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,51200,0.0821217762099372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,10240,0.017887110511461895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,16384,0.019529778096410964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,12288,0.01586933268441094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,16384,0.029607110553317603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,8192,0.018037334084510803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,1024,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,12288,0.02369511127471924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,10240,0.014013333453072442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,7168,0.012562666502263812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,6144,0.01111822244193819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,10240,0.020625778370433383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,8192,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,7168,0.010823999842007955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,8192,0.0177502218219969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,6144,0.010056888891590966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,7168,0.01608000033431583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,4096,0.009630222287442949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,6144,0.01440711153878106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,5120,0.009328888522254096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,5120,0.012319111161761813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,4096,0.008607111043400234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,51200,0.06399199697706434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,3072,0.007870222131411234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,3584,0.008245333201355403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,3584,0.010087110930018956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,2560,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,3072,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,3072,0.009296000003814697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,2048,0.0061048890153567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,2560,0.0075493330756823225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,2560,0.008661333057615492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,1536,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,2048,0.007080000307824876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,3072,768,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,2048,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,1024,0.004187555362780889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,5120,0.009912889036867354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,1536,0.006782222125265334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,768,0.003842666745185852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,1536,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,512,0.003232888877391815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,1024,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,768,0.006305777778228124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,256,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,3584,0.0097884444726838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,768,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,512,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,128,0.0021359999146726397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,512,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,4096,0.010752889017264048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,64,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,256,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2560,32,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,128,0.005865777946180767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,256,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,65536,0.0653182201915317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,128,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,51200,0.05232355660862393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,64,0.00563466673096021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2560,32,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,65536,0.060031996832953564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,51200,0.04737155636151632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,12288,0.01700355609258016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,51200,0.050847109821107656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,16384,0.01995822290579478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,10240,0.017128888103697035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,12288,0.015492444237073263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,12288,0.016233777006467182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2560,1024,0.0063973334100511335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,8192,0.013449778159459433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,10240,0.013175110850069257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,7168,0.012518222133318583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,10240,0.01462577780087789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,8192,0.011417778001891242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,6144,0.01070577816830741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,8192,0.012736000120639801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,5120,0.00960622231165568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,7168,0.011715555356608497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,6144,0.010060444474220276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,6144,0.010040889183680216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,4096,0.008214222060309516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,5120,0.009308444129096137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,3584,0.00758755538198683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,5120,0.00904799997806549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,16384,0.02132266594303979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,4096,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,3072,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,65536,0.06473155816396077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,3584,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,2560,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,16384,0.019197333190176222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,3584,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,2048,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,2560,0.007485333416197035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,3072,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,1536,0.004519999855094486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,2560,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,1024,0.003548444559176763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,2048,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,1536,0.006650666809744305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,768,0.0031591111587153543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,1536,0.006302222195598815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,1024,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,512,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,7168,0.01070844464831882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,768,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,768,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,256,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,512,0.00598933340774642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,512,0.005495999836259418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,256,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,256,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,64,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,4096,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,128,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,32,0.0021537778278191886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,3072,0.007878222399287755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,64,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,65536,0.047453333934148155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,2048,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,32,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,51200,0.03933599922392104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,65536,0.05845866600672404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,16384,0.016215999921162922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,65536,0.0616204473707411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,51200,0.04638222191068861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,2048,1024,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,51200,0.04884355598025852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,12288,0.01590844492117564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,16384,0.018228444788191054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,10240,0.014085332552591959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,12288,0.014450665977266101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,8192,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,2048,128,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,12288,0.015849777393870883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,7168,0.011511111425028907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,10240,0.014353776971499125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,2048,128,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,8192,0.011395555403497485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,6144,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,8192,0.011791110866599612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,5120,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,7168,0.011201777391963534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,6144,0.010010666317409938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,4096,0.007838221887747446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,6144,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,5120,0.00904088881280687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,3584,0.007088888850477006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,5120,0.009072889056470659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,4096,0.008546666966544257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,3072,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,4096,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,3584,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,2560,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,3072,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,3072,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,2048,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,16384,0.01959111127588484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,2560,0.007383999725182851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,1536,0.004607999904288186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,2560,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,2048,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,1024,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,2048,0.00664444433318244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,1536,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,768,0.0032222221295038858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,7168,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,1536,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,512,0.0031662223239739737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,1024,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,256,0.0028328889360030494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,768,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,768,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,512,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,128,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,256,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,512,0.0052888890107472735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,64,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1536,32,0.002128888956374592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,256,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,128,0.005671111245950063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,65536,0.03494666682349311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,3584,0.008299555215570662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,128,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,51200,0.02836266822285122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,64,0.005155555489990446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,16384,0.013814222481515674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,65536,0.05748711029688517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,65536,0.05945689148373074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,51200,0.04621155394448174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,51200,0.04661866691377428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,16384,0.016757332616382174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1536,10240,0.012784889174832238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,10240,0.012319111161761813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,16384,0.019142222073343065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,12288,0.014090667168299357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,8192,0.010503110786279043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,12288,0.015113777584499784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,7168,0.009865778187910715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,10240,0.01316088851955202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,8192,0.011095110740926532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,6144,0.00926755534278022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,8192,0.011701333026091257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,7168,0.010625777973069085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,7168,0.011003555523024665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,5120,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,6144,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,4096,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,6144,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,5120,0.00906755526860555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,3584,0.007183111376232571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,5120,0.008952000074916417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,4096,0.008469333251317343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,4096,0.008175111479229396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,32,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,3584,0.008093333078755273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,3584,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,2560,0.0058648888435628675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,3072,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,3072,0.007438222567240397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,2048,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,2560,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,2560,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,10240,0.012836444709036084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,1536,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,2048,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,1024,0.0032631111227803757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,2048,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1536,1024,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,1536,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,768,0.0035502223504914176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,1024,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,512,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,768,0.005993777679072485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,768,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,256,0.002472888885272874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,512,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,512,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,128,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,256,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,12288,0.0119982221060329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,1024,128,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,32,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,128,0.004950222041871813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,65536,0.028262221150928076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,64,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,32,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,51200,0.023733332753181458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,65536,0.057886222998301186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,65536,0.057095110416412354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,16384,0.014322666658295525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,1536,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,51200,0.04562933246294657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,1024,1024,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,51200,0.04587733414438036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,12288,0.011941333611806234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,16384,0.016805332567956712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,16384,0.01814933286772834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,12288,0.014023999373118082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,8192,0.010070222119490305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,3072,0.007672888537247975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,10240,0.012728888955381183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,10240,0.013144888811641268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,7168,0.009535999761687385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,1024,64,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,6144,0.009010666774378883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,8192,0.01146577795346578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,7168,0.010409778190983666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,5120,0.008193777667151557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,7168,0.010758221977286868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,6144,0.009810667071077559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,4096,0.008767111433876885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,5120,0.009139555195967356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,6144,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,3584,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,5120,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,4096,0.008608889248636033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,3072,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,3584,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,4096,0.008014221986134848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,2560,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,3072,0.007663110891977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,3584,0.00792266676823298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,2048,0.004537777768241035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,3072,0.0074133334888352295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,2560,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,1536,0.0041777777175108595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,10240,0.011063999599880643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,2048,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,2560,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,2048,0.00656444455186526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,1536,0.006715555571847492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,768,0.0031759999692440033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,12288,0.01458044515715705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,1024,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,512,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,1536,0.005988444305128521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,256,0.002521777732504739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,1024,0.005672889037264719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,8192,0.01131377783086565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,512,0.005859555469618903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,128,0.0022746666024128595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,768,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,256,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,512,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,32,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,256,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,128,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,65536,0.024380443824662104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,128,0.005070222334729301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,51200,0.01963733302222358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,64,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,768,32,0.004977777600288391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,65536,0.056735111607445605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,16384,0.010709332923094431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,65536,0.05675377448399862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,51200,0.04388622111744351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,12288,0.009463110731707679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,51200,0.045273777511384755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,16384,0.016618667377365958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,16384,0.017477333545684814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,10240,0.011378666592968835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,1024,0.00350577798154619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,12288,0.0140364451540841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,8192,0.010662222074137794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,12288,0.014728888869285583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,10240,0.01274222218328052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,7168,0.00902844468752543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,8192,0.011008888483047485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,10240,0.01276266657643848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,6144,0.008958222137557136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,768,768,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,5120,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,6144,0.009698666632175446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,768,64,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,4096,0.0068853331936730286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,6144,0.009712888962692684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,5120,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,5120,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,3584,0.006219555520349079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,4096,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,3072,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,4096,0.008079999850855933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,3584,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,2560,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,3072,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,3072,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,2048,0.004782222211360931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,2560,0.00740533322095871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,2560,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,1536,0.004184000194072723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,2048,0.007119999991522894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,2048,0.006346666564544042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,1024,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,1536,0.006745777610275481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,7168,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,1536,0.006123555617200003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,768,0.0033111110743549135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,7168,0.010586666564146677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,1024,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,512,0.002595555658141772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,768,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,1024,0.005767111149099138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,256,0.002424888933698336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,768,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,512,0.005966222120655908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,512,0.005296000176005893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,256,0.004980444494220945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,3584,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,64,0.001992000059949027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,128,0.005660444498062134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,32,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,128,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,65536,0.018000000052981906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,51200,0.015223110715548197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,64,0.005044444567627377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,32,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,16384,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,65536,0.05577244361241659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,51200,0.040023999081717596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,12288,0.008580444587601556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,51200,0.04451377855406868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,16384,0.016546666622161865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,10240,0.008014221986134848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,12288,0.013778666655222574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,12288,0.01443733274936676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,8192,0.007251555720965068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,10240,0.013087111214796701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,10240,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,512,256,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,8192,0.01108000013563368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,8192,0.011335111326641507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,6144,0.009295110901196798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,7168,0.010617777705192566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,7168,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,5120,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,6144,0.009825777676370408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,6144,0.009520889156394536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,4096,0.00684799998998642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,5120,0.009091555244392818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,512,8192,0.011377777490350934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,4096,0.008489777644475302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,4096,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,3584,0.00811644477976693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,3072,0.005528000079923206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,512,128,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,7168,0.008373333348168267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,2560,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,3072,0.007592000067234039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,3072,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,2048,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,2560,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,2560,0.006918222126033571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,1536,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,2048,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,65536,0.04961511161592272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,2048,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,1024,0.003149333306484752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,1536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,1536,0.006157333238257303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,768,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,1024,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,3584,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,1024,0.005684444473849402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,512,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,768,0.005971555494599872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,256,0.002222222172551685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,512,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,768,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,128,0.0021155555215146807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,3584,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,512,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,256,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,256,128,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,256,0.004972444640265571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,32,0.0018639999131361644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,128,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,64,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,32,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,51200,0.01682844426896837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,65536,0.048991110589769155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,65536,0.05327999922964308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,16384,0.013800000150998434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,5120,0.008645333349704742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,51200,0.03939466675122579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,12288,0.010678222609890832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,51200,0.042846222718556724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,256,16384,0.01719555589887831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,10240,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,16384,0.016880000631014507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,12288,0.013783111340469785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,8192,0.008316444853941599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,12288,0.01384888920519087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,10240,0.012475555141766867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,7168,0.00797155582242542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,256,64,0.001793777777089013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,8192,0.011014222270912595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,6144,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,8192,0.011076444553004371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,7168,0.010657777388890585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,65536,0.01534222231970893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,7168,0.010076444182131026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,5120,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,4096,0.006568888823191325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,6144,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,6144,0.009268444445398118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,5120,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,3584,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,4096,0.008358221915033128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,5120,0.00869599978129069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,4096,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,3584,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,2560,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,3584,0.007320000065697565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,3072,0.007864889171388414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,2048,0.004212444441186057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,16384,0.016525333126386006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,2560,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,1536,0.003524444583389494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,2048,0.006973333656787872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,2560,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,1024,0.0031911111954185697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,1536,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,2048,0.0063457778758472866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,768,0.0026586666289303037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,1536,0.005869333528810077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,1024,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,10240,0.012824888858530255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,768,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,1024,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,256,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,768,0.005461333526505365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,512,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,128,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,256,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,512,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,64,0.0018560000591807896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,32,0.0017768889665603638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,193,128,128,0.005882666756709416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,65536,0.024023110667864483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,128,0.004965333475006951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,51200,0.02032444377740224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,3072,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,64,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,16384,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,12288,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,65536,0.052822222312291466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,10240,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,3072,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,51200,0.042747557163238525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,16384,0.016907556189431083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,8192,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,7168,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,6144,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,10240,0.01273333364062839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,8192,0.011173333558771344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,7168,0.010337777435779572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,5120,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,4096,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,6144,0.009379555781682333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,3584,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,5120,0.00869955536392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,3072,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,256,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,2560,0.004798222333192825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,128,32,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,3072,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,2048,0.0042071110672420925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,2560,0.006654222392373615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,1536,0.003504000190231535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,1024,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,12288,0.014058666096793281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,2048,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,768,0.0028000000036425064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,1536,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,512,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,256,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,1024,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,128,0.0019484443797005548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,128,512,0.002459555657373534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,768,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,512,0.005375111268626318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,32,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,256,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,128,0.004601777841647466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,65536,0.026013332936498854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,51200,0.02238044473859999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,64,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,4096,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,16384,0.012441777520709567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,3584,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,12288,0.010302222437328762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,10240,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,51200,0.04233155647913615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,8192,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,16384,0.017052445146772597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,7168,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,12288,0.013912000589900546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,6144,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,10240,0.012726222475369772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,5120,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,4096,0.005655111124118169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,8192,0.011047110789351992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,3584,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,7168,0.01038933296998342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,3072,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,6144,0.009479111267460717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,5120,0.00871733327706655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,2560,0.004863111095296012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,4096,0.007672888537247975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,64,64,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,1536,0.0035137778355015647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,3072,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,1024,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,2560,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,2048,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,768,0.002554666664865282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,512,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,1536,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,64,32,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,256,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,128,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,1024,0.005640888793600931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,65536,0.052667554881837636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,64,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,768,0.005462222215202119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,32,0.001814222170246972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,512,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,256,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,128,0.004860444615284602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,64,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,32,0.004581333448489507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,193,32,2048,0.004184000194072723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,193,32,3584,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,16384,0.28436533610026044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,51200,0.8583120240105523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,12288,0.2986480130089654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,51200,1.2287493811713326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,16384,0.1383777724372016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,10240,0.24254133966233996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,51200,0.41718223359849715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,12288,0.20753955841064453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,16384,0.39291734165615505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,12288,0.10281866788864136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,10240,0.17245422469245064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,8192,0.1992408831914266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,10240,0.08981155686908299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,8192,0.14392088519202337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,7168,0.17338755395677355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,16384,0.3060248957739936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,8192,0.07176799906624688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,12288,0.23354223039415148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,7168,0.12330755922529434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,6144,0.1560764445198907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,7168,0.06446666849984063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,10240,0.1951751046710544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,6144,0.10951733589172363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,8192,0.15663999981350368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,5120,0.1310302284028795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,6144,0.05571022298600939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,4096,0.10701511303583781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,5120,0.09168266587787205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,5120,0.04798666636149088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,6144,0.11708799997965495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,4096,0.07487289110819499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,3584,0.09169511000315349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,4096,0.04025777843263414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,4096,0.0841182205412123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,3584,0.0650728874736362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,3584,0.035926222801208496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,3072,0.07808178000979953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,3584,0.07338844405280219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,2560,0.06609421968460083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,3072,0.03215377860599094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,3072,0.06581955485873751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,2048,0.052904887331856616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,7168,0.1417093276977539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,2560,0.027786665492587622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,2560,0.05792355537414551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,2048,0.0244835548930698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,2048,0.0414284434583452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,1536,0.04086577892303467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,2048,0.049878223074807056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,1024,0.03000977635383606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,1536,0.019292443990707397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,1536,0.03291644321547614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,1536,0.04195644458134969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,1024,0.015545777148670621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,768,0.022950222094853718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,1024,0.025814221964942083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,1024,0.03427555494838291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,3072,0.056177775065104164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,512,0.016511110795868766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,768,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,768,0.021019554800457422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,256,0.010321777727868822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,768,0.0299271113342709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,2560,0.049004445473353066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,512,0.01756177842617035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,512,0.013268444273206921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,51200,0.9272506501939562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,128,0.009152889251708984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,512,0.025651555922296312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,256,0.011734222372372946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,256,0.015656888484954834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,64,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,5120,0.10135022136900161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,128,0.011703111231327057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,65536,32,0.007585778004593319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,256,0.020941333638297185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,128,0.015072888798183866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,128,0.02260355485810174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,64,0.01483022173245748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,51200,0.8504480255974664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,65536,1.0809119542439778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,65536,0.8657849099900988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,51200,0.6851778030395508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,65536,32,0.01550222271018558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,16384,0.27793065706888836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,16384,0.2187804381052653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,12288,0.20986666944291857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,16384,0.11083466476864284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,51200,0.3312568929460314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,12288,0.1610871156056722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,12288,0.08374489016003078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,10240,0.2098862197664049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,10240,0.13713067107730442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,16384,0.24590844578213164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,65536,0.4173751142289903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,8192,0.16159466902414957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,12288,0.18739377127753365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,10240,0.07061600022845797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,8192,0.0572213331858317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,10240,0.15941511260138616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,8192,0.11334044403500027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,7168,0.10044266780217488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,6144,0.1212657822502984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,7168,0.05133777856826782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,8192,0.1289422247144911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,5120,0.1082106696234809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,6144,0.044954667488733925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,6144,0.08876444233788384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,7168,0.11259467071957058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,4096,0.0849377777841356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,5120,0.03871377640300327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,5120,0.07411199808120728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,6144,0.09738400247361924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,3584,0.0740613341331482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,7168,0.14125155078040227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,4096,0.05844177802403768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,4096,0.07024533218807645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,3584,0.029600001043743555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,3072,0.06792266501320733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,3584,0.052073776721954346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,51200,0.7359049055311414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,3072,0.02677511175473531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,3072,0.04604622059398227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,3584,0.06405066781573825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,3072,0.056941330432891846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,2560,0.054013331731160484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,2048,0.04418311185306973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,2560,0.023704888092146978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,2560,0.03974044322967529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,2560,0.050465779172049634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,2048,0.033913777934180364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,1536,0.034676445855034724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,4096,0.03313333458370633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,5120,0.08436089091830784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,2048,0.02010933392577701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,1024,0.025047999289300706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,2048,0.04383822282155355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,1536,0.016519111063745286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,65536,0.9851386811998156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,768,0.019225777851210702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,1536,0.027520888381534155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,1024,0.013807999591032663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,1536,0.037110222710503474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,1024,0.02127200033929613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,512,0.013745778136783175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,768,0.012778666284349231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,256,0.008610666626029545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,512,0.011705777711338468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,768,0.026037333740128413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,512,0.014815110299322339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,128,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,768,0.01720711092154185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,256,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,256,0.013408000270525614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,64,0.005532444351249271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,256,0.018992000155978732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,51200,128,0.0105608892109659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,51200,32,0.00755466686354743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,128,0.015545777148670621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,64,0.012429333395428128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,32,0.013387555877367655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,65536,0.3884764512379964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,65536,0.30841067102220326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,1024,0.030328889687856037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,51200,0.29896177185906303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,65536,0.1430248949262831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,16384,0.10073866446812947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,51200,512,0.022444445225927565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,51200,0.23772001266479492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,16384,0.040489776266945735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,12288,0.07503910859425862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,51200,0.11436000135209824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,16384,0.08747644556893243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,12288,0.06106044186486138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,10240,0.06442399819691975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,12288,0.032165331972969904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,12288,0.06919733020994398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,8192,0.05208977725770739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,10240,0.05216888917817009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,10240,0.0277146663930681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,10240,0.06109688679377238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,7168,0.04644711150063408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,8192,0.02366488840844896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,8192,0.044181333647833936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,65536,0.324111991458469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,51200,0.2593288951449924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,6144,0.04001422060860528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,7168,0.021215111017227173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,7168,0.03846577803293864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,16384,0.08102755414115058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,5120,0.03421688742107815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,6144,0.0337066650390625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,7168,0.0463137792216407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,6144,0.018813333577579923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,4096,0.028351998991436426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,6144,0.043848001294665866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,5120,0.01685688893000285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,5120,0.029264890485339697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,3584,0.025496888491842482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,5120,0.0373271107673645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,4096,0.015327110886573792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,3072,0.022671999202834234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,4096,0.02477600011560652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,4096,0.03225244416130914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,3584,0.013815110756291283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,3584,0.022277333670192297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,2560,0.01865955524974399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,3584,0.029521776570214167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,3072,0.013085333009560903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,2048,0.01516533394654592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,3072,0.020131554868486192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,3072,0.027093332674768236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,2560,0.012087110843923358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,2560,0.017270222306251526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,1536,0.01218933363755544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,2560,0.024420445164044697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,8192,0.051479998562071055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,2048,0.010943999720944298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,2048,0.015094222293959724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,1024,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,51200,128,0.012673777838548025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,2048,0.02203111184967889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,768,0.008103999826643202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,1536,0.012838222086429596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,1536,0.019298666053348117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,1024,0.010662222074137794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,512,0.0063991112013657885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,1024,0.01705777810679542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,768,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,768,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,256,0.004534222185611725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,512,0.0070017774899800616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,512,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,512,0.013743999931547375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,256,0.0068808889223469635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,128,0.0035146665241983202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,256,0.007620444728268518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,256,0.011732444167137146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,64,0.003891555385457145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,128,0.006917333437336816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,128,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,128,0.009640889035330879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,64,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,1536,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,16384,32,0.008044444852405125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,1024,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,65536,0.27870133188035756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,768,0.014930667148696052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,65536,0.22396887673272026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,65536,0.12657511234283447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,51200,0.1726613309648302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,16384,0.07656533188290067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,51200,0.10100711054272121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,16384,0.05919555823008219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,16384,32,0.004222222086456087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,12288,0.05872178077697754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,16384,0.037788444095187716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,16384,0.08584711286756729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,10240,0.048397332429885864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,12288,0.045328888628217906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,12288,0.02983377708329095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,12288,0.0676800012588501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,8192,0.040258665879567467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,10240,0.025732444392310247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,10240,0.03853955533769395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,65536,0.30837512016296387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,51200,0.24393065770467123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,7168,0.035102221700880266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,10240,0.05892622470855713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,8192,0.021856889128684998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,8192,0.033153778976864286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,6144,0.02975911233160231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,8192,0.049908445941077344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,7168,0.028928889168633357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,5120,0.025877333349651758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,7168,0.04551644457711113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,6144,0.01797422269980113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,6144,0.025751999682850305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,4096,0.022142221530278523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,6144,0.040786667002571955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,5120,0.016128889388508268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,5120,0.021980444590250652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,5120,0.035916444328096174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,4096,0.018650665879249573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,4096,0.014465777410401238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,3072,0.016703999704784818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,4096,0.031009776724709406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,3584,0.01700177788734436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,3584,0.013407111167907715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,2560,0.013691555294725629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,3584,0.02825688984658983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,3072,0.012465777496496836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,3072,0.015352000792821249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,51200,0.21566577752431235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,2048,0.01183111137813992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,3072,0.026362667481104534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,2560,0.011600889265537262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,2560,0.013797333670987023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,1536,0.009837333526876237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,7168,0.019876443677478366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,2048,0.00981244444847107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,1024,0.008035555481910706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,2048,0.021399110555648804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,2048,0.012599999705950419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,1536,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,1536,0.01069155583779017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,1536,0.019183110859658983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,1024,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,1024,0.008711111214425828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,512,0.005192888693677055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,3584,0.019130667050679524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,1024,0.01571733256181081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,768,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,768,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,256,0.0038373333712418876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,768,0.014078223043017916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,512,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,128,0.0031795555518733132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,512,0.012778666284349231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,256,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,256,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,256,0.011343999869293638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,64,0.0028728888266616394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,128,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,128,0.006775110960006714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,128,0.009638222555319468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,2560,0.023571555813153584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,64,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,12288,32,0.007078222102589077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,768,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,65536,0.2726462152269152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,51200,0.2104346619711982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,65536,0.2083688841925727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,65536,0.12391289075215657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,512,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,51200,0.1743582222196791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,16384,0.0727377798822191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,51200,0.09852711359659831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,16384,0.03658577799797058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,12288,0.056100441349877246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,16384,0.055903997686174184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,12288,32,0.0031759999692440033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,16384,0.08771999677022298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,10240,0.04502933224042257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,12288,0.02853333287768894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,12288,0.04341333442264133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,8192,0.03801777627733018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,12288,0.06716977887683444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,10240,0.024677332904603746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,65536,0.30616622500949436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,7168,0.033765332566367254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,10240,0.0584133333630032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,51200,0.24169333775838217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,8192,0.020998222960366145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,6144,0.029139555162853662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,8192,0.03157511022355821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,7168,0.027630221512582567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,5120,0.02511288887924618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,6144,0.024081778195169237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,7168,0.044975999328825206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,6144,0.03985155622164408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,4096,0.020862221717834473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,5120,0.015801777442296345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,5120,0.020729778541458976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,5120,0.034880889786614314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,3584,0.018248889181349013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,4096,0.014089778065681458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,4096,0.017973333597183228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,4096,0.030455999904208716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,3072,0.016159999701711867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,3584,0.013003555436929068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,3584,0.016488000750541687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,3584,0.027701333165168762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,2560,0.013731555806265937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,3072,0.012067555553383298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,3072,0.015112888481881885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,3072,0.025407999753952026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,7168,0.019303111566437613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,2048,0.01164533363448249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,2560,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,6144,0.01758399936887953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,2560,0.013808888693650564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,2560,0.022981334063741896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,2048,0.008850666383902231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,2048,0.011917333636018964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,1024,0.007925333248244392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,2048,0.020941333638297185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,1536,0.01034044474363327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,1536,0.018235555953449674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,1024,0.00742222203148736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,768,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,1024,0.014761777387724983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,10240,0.03734577695528666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,768,0.007548444800906711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,768,0.007257777783605788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,512,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,768,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,256,0.004226666771703296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,512,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,8192,0.04943733414014181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,512,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,128,0.0035173334181308746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,512,0.012740444805887012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,1536,0.009687110781669617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,256,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,256,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,64,0.0028568889117903183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,10240,32,0.003216888962520493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,256,0.011330666641394297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,128,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,128,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,1536,0.008074666890833113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,64,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,65536,0.20342755317687988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,32,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,10240,1024,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,51200,0.15174667040506998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,16384,0.05078755484686958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,65536,0.16609156131744385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,65536,0.07704355319341023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,51200,0.1297582255469428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,51200,0.060737775431738965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,16384,0.02404444416364034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,12288,0.04024799995952182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,16384,0.04550844430923462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,16384,0.05190844337145487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,10240,0.033287998702791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,12288,0.019211555520693462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,12288,0.03502488798565335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,51200,0.13725066184997559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,65536,0.173455105887519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,8192,0.028512001037597656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,128,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,10240,0.01688799924320645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,10240,0.030412445465723675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,7168,0.025021332833502028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,10240,0.036870221296946205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,8192,0.01497244503762987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,8192,0.025274667474958632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,6144,0.02126311096880171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,8192,0.03215377860599094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,7168,0.013783999615245394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,7168,0.022439110610220168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,5120,0.018384888768196106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,7168,0.029494222667482164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,6144,0.01277599980433782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,6144,0.019710222880045574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,4096,0.015377778146002026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,6144,0.02632177703910404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,5120,0.011682666838169098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,3584,0.013421333498424955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,5120,0.017315554949972365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,4096,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,4096,0.015171556009186639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,3072,0.01203644441233741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,4096,0.021328889661365088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,3584,0.009904888768990835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,3584,0.014081777797804939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,2560,0.010320888625250923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,12288,0.04206933246718513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,3072,0.008711111214425828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,3072,0.012949333422713809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,2048,0.00923822240697013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,3072,0.01867644488811493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,2560,0.00815466625822915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,2560,0.011712888876597086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,1536,0.00758133331934611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,2560,0.017330666383107502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,2048,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,2048,0.01053244454993142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,1024,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,2048,0.01593777702914344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,1536,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,1536,0.00870488915178511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,768,0.0046844445168972015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,1536,0.014608888162506951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,1024,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,512,0.0035208890007601846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,5120,0.023801777097913954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,1024,0.012298666768603854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,768,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,768,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,256,0.0031991110493739447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,512,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,512,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,128,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,3584,0.01993955506218804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,512,0.011172444456153445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,256,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,256,0.010676444404655032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,64,0.002895999906791581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,8192,32,0.0031662223239739737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,128,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,128,0.006029333505365584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,128,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,64,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,32,0.006304889089531369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,51200,0.1415191094080607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,65536,0.16057333681318495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,1024,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,65536,0.07221688826878865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,16384,0.048716445763905845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,51200,0.05860266420576307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,51200,0.12520621882544622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,768,0.011406222151385413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,12288,0.03698844379848904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,16384,0.04403733213742574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,8192,256,0.006177777631415262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,16384,0.023027555810080633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,65536,0.16109955310821533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,10240,0.03184533450338576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,16384,0.05039022366205851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,12288,0.01845066746075948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,51200,0.12873955567677817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,12288,0.034070223569869995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,8192,0.026425777210129633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,12288,0.04112889038191901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,7168,0.023157333334287006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,10240,0.029496888319651287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,8192,0.014434667097197639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,10240,0.03646577729119195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,8192,0.02470933397610982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,6144,0.019926221834288705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,8192,0.031606223848130964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,7168,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,7168,0.02163822286658817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,5120,0.017198221551047433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,7168,0.02881511052449544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,6144,0.012490666574902005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,6144,0.019367110398080613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,6144,0.025642666551801894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,5120,0.011284444067213269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,4096,0.014793778459231058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,5120,0.017137777474191453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,3584,0.012702222499582501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,5120,0.023333332604832117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,4096,0.00998044427898195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,4096,0.015119999647140503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,3072,0.01143200033240848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,4096,0.02119111186928219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,3584,0.019739554988013375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,65536,0.17886222733391655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,2560,0.010318222145239512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,3072,0.012734221915404001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,3072,0.008631111019187504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,3072,0.018530666828155518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,2048,0.008438222110271454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,2560,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,2560,0.011725333001878528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,2560,0.017095110482639737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,1536,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,2048,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,2048,0.015844444433848064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,1536,0.008833777573373582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,1536,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,1024,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,1536,0.01350488927629259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,3584,0.009333333207501305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,768,0.004810666872395409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,1024,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,1024,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,10240,0.01645866698688931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,512,0.0036720000207424164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,768,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,768,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,768,0.01145600030819575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,256,0.0031582222630580268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,512,0.006223111102978389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,512,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,128,0.0028168888141711554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,512,0.010673777924643623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,2048,0.010392888552612728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,256,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,64,0.0026177778426143858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,256,0.005967999911970562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,256,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,128,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,128,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,128,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,65536,0.1700915495554606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,64,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,32,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,7168,3584,0.013893332746293811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,51200,0.13438310888078478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,65536,0.12139199839697944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,1024,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,65536,0.0698631140920851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,16384,0.043199112017949425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,51200,0.09495021899541219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,51200,0.056877334912618004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,16384,0.02187999917401208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,16384,0.034536888202031456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,16384,0.04776533444722494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,51200,0.12007821930779351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,10240,0.02938755684428745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,12288,0.01779111060831282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,12288,0.02715822226471371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,8192,0.024111110303137038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,7168,32,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,10240,0.015964445140626695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,10240,0.022699554761250813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,7168,0.021225778592957392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,10240,0.03436977664629618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,8192,0.01980266637272305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,8192,0.014185777968830533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,6144,0.01812000075976054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,8192,0.029713776376512315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,7168,0.01310400002532535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,7168,0.01755288905567593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,5120,0.015684444043371413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,7168,0.026741334133678015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,6144,0.012050666742854647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,6144,0.015832000308566622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,4096,0.012820444173283048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,6144,0.024567110670937434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,12288,0.03419288992881775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,5120,0.011049778097205691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,3584,0.011439111497667102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,5120,0.014125333891974555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,5120,0.0222053329149882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,4096,0.00958933350112703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,4096,0.012438221938080259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,3072,0.010323555933104621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,3584,0.008952889177534316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,4096,0.020211555891566806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,3584,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,2560,0.009112888740168678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,3584,0.01866044435236189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,12288,0.03869600097338358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,3072,0.008195555872387355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,2048,0.007808888951937358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,2560,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,3072,0.017662222186724346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,2560,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,1536,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,2560,0.016514667206340365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,2048,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,2048,0.007368000017272101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,1024,0.0053582220441765254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,2048,0.014500444134076437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,1536,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,1536,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,768,0.0042239998777707415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,1536,0.013068444199032254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,1024,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,1024,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,512,0.003535110917356279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,1024,0.011708444191349877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,768,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,768,0.006315555423498154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,256,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,768,0.011100444528791638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,512,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,512,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,128,0.0024986666523747975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,256,0.005671111245950063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,3072,0.010778666370444827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,256,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,64,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,128,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,128,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,6144,32,0.002536000063021978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,128,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,65536,0.1373857789569431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,64,0.005643555687533484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,6144,32,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,65536,0.11966666910383437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,51200,0.11982044908735488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,65536,0.06753955284754436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,16384,0.045900444189707436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,65536,0.1514693366156684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,51200,0.05424177646636963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,65536,0.1497173309326172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,16384,0.020956445071432326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,16384,0.03342399994532267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,12288,0.03509155578083462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,51200,0.11999821662902832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,256,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,10240,0.029583109749688044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,16384,0.0474275549252828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,12288,0.017367111312018502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,8192,0.02458488941192627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,12288,0.025613332788149517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,10240,0.015503111812803479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,10240,0.02199555602338579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,12288,0.0380951099925571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,7168,0.02160088883505927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,10240,0.033352000845803156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,8192,0.013743110828929476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,6144,0.01871555546919505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,8192,0.01959288948112064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,7168,0.01291822228166792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,7168,0.017165333032608032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,5120,0.015819556183285184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,7168,0.026319111386934917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,6144,0.011715555356608497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,6144,0.015615110596021017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,51200,0.09335644377602471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,6144,0.02423022190729777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,5120,0.010384889112578498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,5120,0.013788444300492605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,3584,0.011811555259757571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,5120,0.021680000755521987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,4096,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,512,0.010804444551467896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,3072,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,4096,0.019848000672128465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,3584,0.011406222151385413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,3584,0.01849688920709822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,2560,0.008954666554927826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,3072,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,3072,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,3072,0.01719199948840671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,2048,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,2560,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,2560,0.009642666412724389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,8192,0.028660443094041612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,1536,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,2560,0.01585422290696038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,2048,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,4096,0.013632888595263163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,2048,0.014096000128322177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,1536,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,1536,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,1024,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,1536,0.012606222596433429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,768,0.003965333518054751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,1024,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,4096,0.012318222059143914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,1024,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,512,0.003531555748648114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,3584,0.008368000388145447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,1024,0.011391110718250275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,768,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,256,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,768,0.011417778001891242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,512,0.006225777582989798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,512,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,128,0.002529777793420686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,512,0.010703999963071613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,256,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,256,0.010000000397364298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,128,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,64,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,5120,32,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,2048,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,128,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,65536,0.11070933606889512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,64,0.005640000104904175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,32,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,51200,0.08340444167455037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,65536,0.06504088640213013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,768,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,65536,0.14863644705878362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,51200,0.08728266424602932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,51200,0.051757332351472646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,16384,0.03398222062322829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,256,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,51200,0.11818488438924153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,12288,0.029488888051774766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,16384,0.020208888583713107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,16384,0.031565331750445895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,5120,128,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,16384,0.0462719996770223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,12288,0.016843555702103507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,12288,0.024663110574086506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,8192,0.020530665914217632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,12288,0.036944889360004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,10240,0.01499555508295695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,10240,0.02196088929971059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,7168,0.01777333352300856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,8192,0.013100444442696042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,10240,0.03257066673702664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,65536,0.11295910676320393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,8192,0.01870844430393643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,6144,0.015787555111779105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,8192,0.02812533246146308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,7168,0.012094222009181976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,5120,0.013401777380042605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,6144,0.010764444039927589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,7168,0.025638222694396973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,6144,0.015153777268197803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,4096,0.01163200040658315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,6144,0.02346399923165639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,5120,0.009535111486911774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,3584,0.008591111335489485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,5120,0.021546666820844013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,4096,0.008773333496517604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,4096,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,3072,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,4096,0.019056000643306308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,3584,0.008357333640257517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,10240,0.02420177724626329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,2560,0.006748444504208035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,3584,0.017892445127169292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,3072,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,3072,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,2048,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,3072,0.016840888394249808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,2560,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,2560,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,1536,0.004973333328962326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,2560,0.015674667225943673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,2048,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,7168,0.01684533390733931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,2048,0.013769778112570444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,1024,0.003891555385457145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,1536,0.0069066666894488865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,1536,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,768,0.0032720000793536506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,1536,0.012157333393891653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,5120,0.013491555220550962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,1024,0.006542222367392645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,1024,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,512,0.002970666728085942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,1024,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,768,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,768,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,256,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,768,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,3584,0.010832889212502373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,512,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,512,0.010045333041085137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,256,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,256,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,256,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,64,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,128,0.0052933332820733385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,128,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,32,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,128,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,2048,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,65536,0.10898844401041667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,32,0.0058497778243488735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,51200,0.08672089046902126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,65536,0.10785600211885239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,65536,0.06386666827731662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,65536,0.14832622475094265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,51200,0.050698667764663696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,512,0.005980444451173146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,16384,0.03825689024395413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,51200,0.08518489201863606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,4096,128,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,12288,0.03052799900372823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,51200,0.11790311336517334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,16384,0.01993955506218804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,10240,0.02179199953873952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,12288,0.01660711069901784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,16384,0.045831110742357045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,12288,0.024597333537207708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,8192,0.018287110659811232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,12288,0.03666399916013082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,10240,0.020943111843532983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,10240,0.014859555496109856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,4096,64,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,7168,0.015818667080667283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,8192,0.012973333398501078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,10240,0.03201244605912103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,6144,0.014039110806253223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,8192,0.018295110927687753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,7168,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,8192,0.027959111664030287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,7168,0.016584000653690763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,6144,0.010692444112565784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,7168,0.025973333252800837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,6144,0.01482222146458096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,4096,0.010612444745169746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,6144,0.023694222172101338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,5120,0.013402666482660504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,3584,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,5120,0.021293333835071985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,4096,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,4096,0.01926399932967292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,16384,0.030321776866912842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,3072,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,3584,0.008290666672918532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,3584,0.010300444232092964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,2560,0.00656088896923595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,3584,0.017247110605239868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,3072,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,3072,0.009327111144860586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,2048,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,2560,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,2560,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,2560,0.014856888188256158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,1536,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,5120,0.012245333029164208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,2048,0.007272889216740926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,2048,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,1024,0.0036391110883818734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,2048,0.01351377781894472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,1536,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,1536,0.007073777417341868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,768,0.0033528889632887314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,1536,0.012436444560686747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,1024,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,1024,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,512,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,4096,0.011199110911952125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,1024,0.011035555766688453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,768,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,768,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,256,0.00258933338854048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,768,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,512,0.005850666513045629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,128,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,512,0.00998222248421775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,256,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,64,0.0021840000732077495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,256,0.010126222338941362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,128,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3584,32,0.0022693334354294673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,128,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,128,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,65536,0.09668444262610541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,64,0.005640888793600931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,32,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,51200,0.07735377550125122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,65536,0.0702586637602912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,65536,0.06286578045950995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,5120,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,16384,0.028203556934992473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,51200,0.05496266815397474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,51200,0.04974666568968031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,51200,0.11839289135403103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,16384,0.01963733302222358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,16384,0.02095555596881443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,12288,0.019863999552196927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,16384,0.045307556788126625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,12288,0.016410667035314772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,10240,0.016528889536857605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3584,256,0.005623111294375525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,12288,0.016910221841600206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,3072,0.016072000066439312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,8192,0.014304000470373364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,12288,0.036567111810048424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,10240,0.014860444598727755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,7168,0.012695999609099494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,10240,0.015228443675571017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,8192,0.012438221938080259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,8192,0.013386666774749756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,6144,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,8192,0.027748444014125403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,7168,0.012400888734393649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,7168,0.02545333405335744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,6144,0.010428444378905825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,6144,0.011232888533009423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,65536,0.147544887330797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,6144,0.023052444060643513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,4096,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,5120,0.009324444664849175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,5120,0.010068444742096795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,5120,0.021366222037209406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,3584,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,4096,0.008612444831265343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,4096,0.018906666172875296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,3584,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,3072,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,3584,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,3584,0.017281777328915067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,10240,0.031862222486072116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,2560,0.00645955569214291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,3072,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,3072,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,2048,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,3072,0.015499555402331881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,2560,0.007154666715198093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,5120,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,1536,0.004422222160630756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,2048,0.007078222102589077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,2560,0.015030221806632148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,2048,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,1024,0.00351555562681622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,2048,0.01346666696998808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,1536,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,1536,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,768,0.003177777760558658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,4096,0.00868888861603207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,1536,0.01239111108912362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,1024,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,1024,0.0057946667075157166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,512,0.002814222127199173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,1024,0.010787555740939246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,768,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,768,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,256,0.002487999904486868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,512,0.00553688903649648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,512,0.010021333065297868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,128,0.0021991110924217436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,256,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,256,0.0057297779454125305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,256,0.009303111169073317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,64,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,2560,0.0075911109646161394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,3072,32,0.002222222172551685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,128,0.00498577786816491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,65536,0.07253333595063952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,128,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,64,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,3072,32,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,51200,0.05700800153944227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,65536,0.06146044201321072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,65536,0.06749422020382352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,65536,0.14818310737609863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,51200,0.04868266648716397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,16384,0.023802666200531855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,51200,0.05282577872276306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,768,0.010515555739402771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,12288,0.02238666680124071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,51200,0.11733244525061713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,16384,0.019621334142155118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,7168,0.011360888679822287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,16384,0.020390222469965618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,10240,0.019328888919618394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,12288,0.015871110889646742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,128,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,16384,0.04525244567129347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,12288,0.01633777717749278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,12288,0.03586311141649882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,10240,0.013793778088357715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,10240,0.03154666556252374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,8192,0.011913778053389655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,7168,0.012253333297040729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,8192,0.027485332555241052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,7168,0.011104889214038849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,6144,0.01109155515829722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,7168,0.02554399934079912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,512,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,6144,0.010724444356229572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,5120,0.009711110757456886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,6144,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,8192,0.013739555246300168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,5120,0.00924622184700436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,10240,0.014915555715560913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,5120,0.020670221911536325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,5120,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,8192,0.013069333301650153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,3584,0.007553777760929531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,4096,0.008624888956546783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,4096,0.008307555483447181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,4096,0.01789155602455139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,7168,0.012054222325483957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,3584,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,3584,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,3584,0.01683999929163191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,2560,0.006224888894293044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,3072,0.007325333025720384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,3072,0.015187554889255099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,6144,0.023178666830062866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,2048,0.005478222337034013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,2560,0.00702311098575592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,2560,0.00755555596616533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,2560,0.014501333236694336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,1536,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,2048,0.007139555282062954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,4096,0.008351999852392409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,2048,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,1024,0.003552000141806073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,2048,0.013454222016864352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,1536,0.006899555524190267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,768,0.0031884445084465873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,1536,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,1024,0.0063795554969045855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,1536,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,512,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,3072,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,1024,0.011057777537239922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,768,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,768,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,256,0.0028533333291610083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,768,0.01034044474363327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,512,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,512,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,128,0.0024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,3072,0.00793688909875022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,256,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,512,0.00999111102686988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,256,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,64,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2560,32,0.002510222295920054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,128,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,128,0.004956444518433677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,128,0.0075804442167282104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,64,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,32,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,51200,0.04699288805325826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,65536,0.059903999169667564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,65536,0.06465511189566718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,16384,0.02074399921629164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,65536,0.14652000533209905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,51200,0.04770400126775106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2560,1024,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,51200,0.11712355083889431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,12288,0.01738666660255856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,16384,0.019262222780121695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,16384,0.0446648895740509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,12288,0.015216888652907478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,10240,0.015218666858143277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,12288,0.016171556380059984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,12288,0.035622222556008235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,8192,0.013735999663670858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,256,0.009381333159075843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,10240,0.014474666780895658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,10240,0.03152533372243246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,7168,0.012196444802814059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,8192,0.011679111255539788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,8192,0.012974222501118978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,8192,0.027496889233589172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,6144,0.010970667004585266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,7168,0.010762666662534079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,7168,0.011733333269755045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,51200,0.05067111055056254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,5120,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,7168,0.02531822191344367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,6144,0.010017777482668558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,6144,0.010545777777830759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,4096,0.008369777765538957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,16384,0.01982133256064521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,6144,0.02292355563905504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,5120,0.009212444225947062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,5120,0.009298666483826106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,3584,0.007595555649863348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,5120,0.019931554794311523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,4096,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,4096,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,3072,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,3584,0.00831822223133511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,10240,0.013447999954223633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,3584,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,65536,0.06156533294253879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,2560,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,3584,0.016523554921150208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,3072,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,3072,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,2048,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,2560,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,3072,0.015722667177518208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,2560,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,1536,0.0042053332759274375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,2560,0.01407911048995124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,2048,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,2048,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,1024,0.0035093331502543557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,2048,0.013162666724787818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,1536,0.006822222222884496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,768,0.003144888828198115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,1536,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,1024,0.006555555595291986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,1024,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,512,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,1024,0.011152889165613385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,768,0.006205333189831839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,768,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,256,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,768,0.010709332923094431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,512,0.0053084443012873335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,4096,0.017695110705163743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,512,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,256,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,256,0.005643555687533484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,256,0.009902222288979424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,128,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,64,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,128,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,128,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,65536,0.04948711064126757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,64,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,32,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,2048,1536,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,51200,0.042005333635542125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,65536,0.05859377649095324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,65536,0.06177155839072334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,512,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,16384,0.015967110792795818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,51200,0.048680888281928174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,51200,0.04652533266279432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,128,0.0022071111533376905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,12288,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,51200,0.11722666687435573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,16384,0.01829600003030565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,16384,0.019430221782790292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,16384,0.044165333112080894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,12288,0.014490667316648694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,2048,32,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,12288,0.015798222687509324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,8192,0.01275022245115704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,12288,0.03529244330194261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,10240,0.013079110946920184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,10240,0.013872000078360239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,7168,0.011648000114493899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,8192,0.011365333365069495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,10240,0.031146665414174397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,8192,0.012063110868136087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,6144,0.010325333310498131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,8192,0.026738666825824316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,7168,0.010680889089902243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,7168,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,5120,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,7168,0.025017778078715008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,6144,0.010211555494202508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,4096,0.007808888951937358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,6144,0.009729777773221334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,6144,0.022063111265500385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,5120,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,3584,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,5120,0.01960088809331258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,4096,0.00852177788813909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,4096,0.008379555410808986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,3072,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,4096,0.01771377854877048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,10240,0.012716444830099741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,3584,0.008045333127180735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,3584,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,2560,0.005447111195988125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,3584,0.016333333320087857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,3072,0.00778311077091429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,2048,0.004553777890072929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,3072,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,2560,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,3072,0.01568888955646091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,2560,0.014132445057233175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,1536,0.0038862224254343244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,2048,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,2048,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,2048,0.013456888496875763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,1024,0.0032675556010670135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,1536,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,1536,0.00601333338353369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,1536,0.011749332977665795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,768,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,5120,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,1024,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,65536,0.14639643828074136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,512,0.00254844439526399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,1024,0.011047110789351992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,768,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,768,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,256,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,768,0.010332444475756751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,512,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,512,0.0052871112194326185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,128,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,512,0.009741333623727163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,256,0.0057591112951437635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,256,0.005099555684460534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,2560,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,64,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,256,0.009327111144860586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1536,32,0.002128888956374592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,128,0.005628444254398346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,128,0.005022222383154763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,65536,0.035671111610200666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,128,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,64,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,51200,0.029723554849624634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1536,32,0.004971555537647671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,65536,0.05717511309517754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,65536,0.059545775254567467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,16384,0.013410666750537025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,51200,0.046105778879589505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,65536,0.1464008887608846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,1024,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,51200,0.04663999875386556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,16384,0.01891377733813392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,16384,0.01685244507259793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,10240,0.010618666807810465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,16384,0.04389155573315091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,12288,0.014104000396198697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,12288,0.03534844517707825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,8192,0.009518221848540837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,10240,0.013085333009560903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,10240,0.03051999873585171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,8192,0.01146044416560067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,7168,0.009123555488056606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,8192,0.02608266638384925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,7168,0.010556444525718689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,6144,0.00833511104186376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,7168,0.024007999234729346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,12288,0.011810666984981961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,51200,0.11593333880106609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,6144,0.009679999616410997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,5120,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,6144,0.00978488889005449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,12288,0.014849777022997538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,6144,0.02172800070709652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,5120,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,5120,0.008735111190213097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,5120,0.019254222512245178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,3584,0.0063031112982167145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,4096,0.008016889293988546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,4096,0.00831288927131229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,4096,0.017246221502621967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,8192,0.011153777440388998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,3072,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,3584,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,3584,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,2560,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,3072,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,3584,0.016505777835845947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,3072,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,7168,0.010669333239396414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,3072,0.015141333142916361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,2560,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,2560,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,1536,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,2560,0.014074666632546319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,2048,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,2048,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,1024,0.0031991110493739447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,2048,0.013088000317414602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,1536,0.0068604445291890045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,4096,0.007585778004593319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,1536,0.006082666830884085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,768,0.002871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,1536,0.012083555261294046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,10240,0.012876444392734103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,1024,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,512,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,1024,0.011016888750924004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,768,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,768,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,256,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,768,0.010384889112578498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,512,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,512,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,512,0.010370666782061258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,256,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,2048,0.004527111020353106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,64,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,128,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,128,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,32,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,65536,0.028900444507598877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,64,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,32,0.004945777770545748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,65536,0.058147556251949735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,51200,0.023527999718983967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,65536,0.056959999932183154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,65536,0.14589867326948378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,51200,0.046205331881841026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,16384,0.011445333560307821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,51200,0.045765333705478244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,51200,0.11553066306644016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,256,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,16384,0.01665511065059238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,16384,0.017995556195576985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,12288,0.010049777726332346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,10240,0.011540444360838996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,128,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,12288,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,12288,0.014583999911944071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,8192,0.01038044442733129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,12288,0.03462222218513489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,10240,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,10240,0.013032000097963544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,7168,0.009533333281675974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,10240,0.030164443784289893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,8192,0.011400000088744693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,1024,1024,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,6144,0.008896888958083259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,8192,0.011349333657158745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,7168,0.010514666636784872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,7168,0.010656000011497073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,5120,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,7168,0.023750222391552393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,6144,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,6144,0.009770666559537252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,4096,0.006923555499977536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,6144,0.021287999219364587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,5120,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,5120,0.01923377811908722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,3584,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,4096,0.008616000413894653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,16384,0.04380977816051907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,4096,0.01724799970785777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,3072,0.005533333453867171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,3584,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,3584,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,3584,0.016140444411171805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,2560,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,3072,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,3072,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,3072,0.014816888504558139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,2048,0.004514666481150521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,2560,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,2560,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,2560,0.01408266690042284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,1536,0.0038382220599386427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,1024,128,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,2048,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,2048,0.012839111189047495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,1024,0.0032071111102898917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,5120,0.008730666504965888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,1536,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,1536,0.006010666489601135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,768,0.0028373334142896864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,4096,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,1536,0.012104888757069906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,1024,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,512,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,1024,0.005649777750174205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,1024,0.011094222466150919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,768,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,256,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,768,0.010296888649463654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,512,0.009710222482681274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,256,0.005035555611054103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,128,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,256,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,8192,0.02607111136118571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,256,0.00903911143541336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,64,0.001814222170246972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,2048,0.0069884442620807225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,128,0.004955555415815777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,768,32,0.0018213333355055915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,65536,0.022982221510675218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,64,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,32,0.005084444251325396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,65536,0.05673422416051229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,65536,0.05681688918007744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,768,0.005535111245181825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,51200,0.01864888932969835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,65536,0.14550666014353433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,512,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,16384,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,51200,0.04461333486768934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,51200,0.04547199938032362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,16384,0.017641777793566387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,51200,0.11585777335696751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,16384,0.016532444291644625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,12288,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,128,0.00564533347884814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,10240,0.009436444275909001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,12288,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,128,0.007335111498832703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,12288,0.014478221535682678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,8192,0.008237332933478886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,12288,0.033987555238935686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,10240,0.012734221915404001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,10240,0.01277599980433782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,7168,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,10240,0.029923554923799302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,8192,0.011362666885058085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,8192,0.02574577762020959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,6144,0.007002666592597961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,7168,0.010483555495738983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,768,512,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,7168,0.010670222342014313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,5120,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,7168,0.023781332704755995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,6144,0.009711999860074785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,6144,0.009529777699046666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,4096,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,5120,0.008982222113344405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,5120,0.00869333330127928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,5120,0.019542222221692402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,16384,0.04310577776696947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,3584,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,4096,0.008469333251317343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,4096,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,3072,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,4096,0.01717511150572035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,3584,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,3584,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,2560,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,3584,0.016507556041081745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,3072,0.007932444413503012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,3072,0.007249777515729268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,8192,0.011045333411958484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,2048,0.004524444540341695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,3072,0.014835556348164877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,2560,0.0074471111098925276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,1536,0.0038337777886125776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,2560,0.006907555378145642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,2048,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,2560,0.014467555615637036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,2048,0.006319111006127463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,1024,0.0031555555760860443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,2048,0.012833777401182385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,1536,0.006132444573773279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,1024,0.006400888992680444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,1536,0.012051555845472546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,1024,0.005662222289376789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,6144,0.020964443683624268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,512,0.00250133333934678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,1024,0.011076444553004371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,768,0.005967999911970562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,256,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,768,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,768,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,512,0.005414222263627582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,128,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,512,0.010206222534179688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,256,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,256,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,64,0.0018595556418100994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,256,0.009734222458468543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,32,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,128,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,65536,0.01925155520439148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,128,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,128,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,64,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,51200,0.01649599936273363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,512,32,0.005103999955786599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,65536,0.05315555466545952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,65536,0.05563377671771579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,1536,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,16384,0.009000889129108852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,512,768,0.0027991111079851785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,65536,0.1453146669599745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,51200,0.03995022177696228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,51200,0.04447022080421448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,12288,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,16384,0.016515556308958266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,10240,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,51200,0.11593243810865615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,12288,0.013780444032616086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,16384,0.04281333420011732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,512,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,12288,0.01423111061255137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,8192,0.006196444647179709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,12288,0.03386133246951633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,10240,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,10240,0.012718222207493253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,7168,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,10240,0.029727998707029555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,8192,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,6144,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,8192,0.011192888849311404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,8192,0.0255031105544832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,7168,0.010673777924643623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,7168,0.010335110955768162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,5120,0.006647999915811751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,7168,0.02334844403796726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,6144,0.009714666340086196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,6144,0.009415999882751042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,4096,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,5120,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,6144,0.021346666746669348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,5120,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,3584,0.005879111174080107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,5120,0.01926488843229082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,4096,0.008347555167145198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,4096,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,3072,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,4096,0.01720622181892395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,3584,0.00812888890504837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,2560,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,16384,0.01720977822939555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,3584,0.016545777519543965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,3072,0.007677333222495184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,3072,0.007048000064161088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,3072,0.014860444598727755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,2560,0.007455999652544658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,2560,0.0068604445291890045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,1536,0.0038524443904558816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,2560,0.013800889253616333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,2048,0.0070159998204973005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,2048,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,1024,0.003149333306484752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,2048,0.012792888614866467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,1536,0.006637333167923822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,768,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,1536,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,1536,0.01181688904762268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,1024,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,512,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,1024,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,768,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,768,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,256,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,768,0.010680889089902243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,512,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,512,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,128,0.0018595556418100994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,512,0.01036266651418474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,3584,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,256,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,256,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,64,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,32,0.001814222170246972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,256,2048,0.004214222232500712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,256,0.009669333696365356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,65536,0.012843555874294706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,128,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,128,0.0075937774446275495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,51200,0.011410666836632622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,65536,0.049124442868762545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,32,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,65536,0.05403822329309252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,16384,0.006072888771692912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,51200,0.03952888978852166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,65536,0.1449484427769979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,51200,0.04307555490069919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,12288,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,1024,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,16384,0.016514667206340365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,51200,0.1148373285929362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,16384,0.017217778497272067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,10240,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,16384,0.04216622312863668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,12288,0.014226666755146451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,8192,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,12288,0.03357422351837158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,10240,0.01275111072593265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,10240,0.0127368892232577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,7168,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,8192,0.011025778121418424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,8192,0.011018666956159802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,8192,0.02539377742343479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,7168,0.01035199976629681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,6144,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,7168,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,256,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,5120,0.005636444522274866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,7168,0.023361777265866596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,6144,0.009975999593734741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,6144,0.009372444616423713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,5120,0.009016888837019602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,6144,0.020946666598320007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,5120,0.00867377801073922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,4096,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,5120,0.01888088881969452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,4096,0.008445333275530074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,3584,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,12288,0.013774221969975365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,4096,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,3072,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,3584,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,3584,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,10240,0.029515554507573444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,3584,0.015973332855436537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,128,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,3072,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,3072,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,2048,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,3072,0.014861333701345654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,2560,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,2560,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,1536,0.003551111039188173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,2560,0.013788444300492605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,2048,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,2048,0.01275288893116845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,1536,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,1024,0.0029564443975687027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,1536,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,1536,0.011737777955002256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,768,0.0028008888992998335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,1024,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,1024,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,1024,0.010785777535703448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,4096,0.01754044493039449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,512,0.0025191110455327565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,768,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,768,0.01035199976629681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,2560,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,256,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,512,0.005789333333571752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,512,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,128,0.0018773333479960759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,512,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,256,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,256,0.00500266667869356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,64,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,256,0.008965333302815756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,128,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,128,32,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,128,0.004968889057636261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,2048,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,51200,0.011003555523024665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,16384,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,64,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,128,32,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,12288,0.0052826665341854095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,10240,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,65536,0.053025778796937734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,8192,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,51200,0.04247200157907274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,7168,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,16384,0.01722311145729489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,12288,0.013806222213639153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,6144,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,10240,0.012453333371215396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,5120,0.004946666873163647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,768,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,4096,0.004553777890072929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,8192,0.011069333387745751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,3584,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,7168,0.01035199976629681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,3072,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,2560,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,6144,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,5120,0.008349333372380998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,2048,0.004570666700601578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,4096,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,3584,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,1024,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,3072,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,768,0.0028355556229750314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,2560,0.006719111154476802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,512,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,65536,0.009296000003814697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,256,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,1536,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,128,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,1024,0.005655999812814925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,64,0.0018151110659042995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,768,0.005374222166008419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,512,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,65536,0.014035556051466199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,256,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,51200,0.011535111400816174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,128,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,64,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,32,0.00500177757607566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,12288,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,65536,0.05275377631187439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,51200,0.04212622178925408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,16384,0.017186666528383892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,12288,0.014127999544143677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,8192,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,64,1536,0.0034968890249729156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,7168,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,10240,0.012777778009573618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,6144,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,64,2048,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,8192,0.011065777805116443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,5120,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,4096,0.004556444370084339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,7168,0.00997511131895913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,3584,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,6144,0.009406222237481011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,3072,0.0046302220887608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,5120,0.008683555656009251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,4096,0.007723555796676212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,3584,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,16384,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,2048,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,1536,0.003485333174467087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,3072,0.0069982219073507525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,1024,0.002893333426780171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,2560,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,768,0.00282577777074443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,2048,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,512,0.0025111111915773815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,1536,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,256,0.0019448887970712448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,1024,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,128,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,768,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,512,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,32,0.0018231111268202465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,256,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,128,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,64,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,192,32,32,0.004726222405831019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,2560,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,16384,0.38463200463189023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,192,32,10240,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,12288,0.2917439937591553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,16384,0.13654222753312853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,51200,0.8492470847235786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,51200,1.205044428507487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,12288,0.20437333318922254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,10240,0.2353093359205458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,8192,0.19450222121344674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,10240,0.16849511199527314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,12288,0.10248711374070908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,51200,0.40109867519802517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,8192,0.13902044296264648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,7168,0.16757955816056994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,6144,0.14364533954196507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,7168,0.12456266085306804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,10240,0.08745866351657444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,7168,0.062409778436024986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,8192,0.07052977879842122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,5120,0.12194577852884929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,4096,0.10164178080028957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,6144,0.05499200026194254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,6144,0.10374310943815444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,5120,0.0922586652967665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,3584,0.0873146653175354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,3072,0.07562577724456787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,4096,0.07181244426303439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,4096,0.039507554637061224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,3584,0.035183111826578774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,2560,0.06273333231608073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,3072,0.03139111068513658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,2048,0.050759110185835093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,2560,0.02789688772625393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,2560,0.04850222335921394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,1536,0.039464887645509504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,2048,0.024002666274706524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,2048,0.04122399952676561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,16384,0.2763644324408637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,1024,0.028891556792789038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,1536,0.01886311173439026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,1536,0.03222577770551046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,768,0.021700445148679946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,1024,0.015470221638679504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,512,0.01568711135122511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,1024,0.02530488868554433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,5120,0.04666399955749512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,768,0.014092443717850579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,256,0.009657777845859528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,768,0.024136889312002394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,512,0.013207111093733044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,128,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,256,0.011723555624485016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,3584,0.06306311157014635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,64,0.006559999866618051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,256,0.015785778562227886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,65536,32,0.006723555425802867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,128,0.014807111687130399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,3072,0.05542666713396708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,64,0.014783110883500842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,32,0.015225778023401896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,65536,512,0.01789422167672051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,65536,1.1457173029581706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,65536,128,0.011634666886594562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,16384,0.29223378499348956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,65536,0.4071439901987712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,51200,0.6745582156711154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,16384,0.21607022815280488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,16384,0.10825600226720174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,65536,0.869847986433241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,51200,0.9043448766072592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,12288,0.22286934322781035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,51200,0.3167671097649468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,12288,0.16044978300730386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,10240,0.20002222061157227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,12288,0.082223998175727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,8192,0.16309332847595215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,7168,0.14424622058868408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,8192,0.05669599771499634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,8192,0.11219822035895453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,10240,0.0685244467523363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,7168,0.09965422418382432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,7168,0.05077777637375725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,10240,0.13412355052100286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,6144,0.12375022305382623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,5120,0.10502488745583428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,4096,0.08641688691245185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,6144,0.04390933447413974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,3584,0.07571911149554782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,5120,0.03806844353675842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,5120,0.07099022467931111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,4096,0.03287022312482198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,3072,0.06597866614659627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,4096,0.05793155564202202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,3584,0.029151999288135107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,3584,0.051707555850346885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,3072,0.026213333010673523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,2048,0.04522933231459724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,3072,0.045363555351893105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,2560,0.023576888773176406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,2560,0.039164443810780845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,2048,0.019648000597953796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,1024,0.026084444589085046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,2048,0.033365332418017916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,1536,0.016410667035314772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,1536,0.02709688908523983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,768,0.019578667150603402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,1024,0.013935110635227628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,512,0.014398222168286642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,1024,0.020883555213610332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,768,0.020690666304694284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,512,0.011728888584507836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,256,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,6144,0.08430044518576728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,768,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,512,0.015166223049163818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,128,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,64,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,256,0.010296000374688042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,51200,128,0.010544000400437249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,256,0.013398222625255585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,2560,0.055919110774993896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,128,0.012486221889654795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,64,0.012736000120639801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,51200,32,0.013487999637921652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,1536,0.035536001125971474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,65536,0.3707146644592285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,51200,0.29263554679022896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,16384,0.09522844685448541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,65536,0.30332088470458984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,65536,0.14107555813259548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,12288,0.07327822181913587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,51200,0.2352008819580078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,10240,0.06166578001446194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,16384,0.07908444272147284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,16384,0.04031555520163642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,12288,0.031621333625581526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,12288,0.06014044417275322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,10240,0.02751111156410641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,51200,32,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,10240,0.051389333274629384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,8192,0.02349511120054457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,7168,0.04488799969355265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,8192,0.04366933306058248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,6144,0.03852977686458164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,5120,0.03327200147840712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,7168,0.03803999887572394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,6144,0.018806222412321303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,4096,0.027631110615200464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,6144,0.03345955411593119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,5120,0.016754666964213055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,3584,0.02444711161984338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,5120,0.02882666720284356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,4096,0.014927999840842353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,3072,0.02182933357026842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,51200,0.1115715503692627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,3584,0.013785777820481194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,2560,0.01792533364560869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,3584,0.02194044490655263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,3072,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,8192,0.051129778226216636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,2560,0.01202133297920227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,3072,0.01959022217326694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,2048,0.014817777607176038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,2560,0.017064889272054035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,2048,0.010879110958841113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,1536,0.011700444751315646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,1024,0.00903911143541336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,2048,0.014899555179807874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,1536,0.012736000120639801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,7168,0.0211217784219318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,768,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,1024,0.008247111406591203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,512,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,1024,0.010687111152542962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,768,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,256,0.004027555386225383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,768,0.009958222508430481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,512,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,128,0.0034924443397257063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,512,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,256,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,4096,0.0244195560614268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,64,0.0031973332580592898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,16384,32,0.0031848889258172777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,256,0.007592889169851939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,128,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,128,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,64,0.007336888876226213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,16384,32,0.008099555969238281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,65536,0.272269328435262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,65536,0.2172222269905938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,51200,0.21501689487033418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,65536,0.12579822540283203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,16384,1536,0.009208888643317753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,16384,0.0711617800924513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,12288,0.056127998563978404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,51200,0.16904266675313315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,51200,0.10064888662762111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,16384,0.037237332926856145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,12288,0.029734220769670274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,12288,0.04452355702718099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,16384,0.05808622307247586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,8192,0.039260443713929914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,10240,0.025615110993385315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,7168,0.034099555677837796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,10240,0.03828977876239353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,8192,0.021715554926130507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,6144,0.029750221305423315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,8192,0.03286666671435038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,7168,0.01978222197956509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,7168,0.02871733241611057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,6144,0.017846221725145977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,6144,0.025358221597141687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,5120,0.016144000821643405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,3584,0.018559111489189994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,5120,0.0218231115076277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,4096,0.014436443646748861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,4096,0.01872977779971229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,3072,0.016565332810084026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,3584,0.013457777599493662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,2560,0.013741333451535968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,3584,0.01685244507259793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,3072,0.012625777886973487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,2048,0.011773332953453064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,10240,0.04613777663972643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,3072,0.015252444479200574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,2560,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,1536,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,2560,0.013678222066826291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,1024,0.007928888830873702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,2048,0.009745777481132084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,1536,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,768,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,1024,0.007762666377756331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,1536,0.010672888822025724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,512,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,1024,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,5120,0.025620443953408137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,256,0.004221333397759331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,4096,0.022272000710169475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,768,0.007452444069915348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,768,0.008655110994974772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,512,0.0069955554273393415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,128,0.003526222374704149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,64,0.003170666595300039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,512,0.007099555598364935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,256,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,12288,32,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,12288,128,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,256,0.006650666809744305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,64,0.00664444433318244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,32,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,65536,0.26158044073316783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,51200,0.20474844508700898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,65536,0.1230471134185791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,16384,0.06888800197177462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,2048,0.012438221938080259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,51200,0.1601839992735121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,12288,0.05329599976539612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,16384,0.0554657777150472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,16384,0.036342223485310875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,10240,0.04529777831501431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,12288,0.02826311190923055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,10240,0.024542222420374554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,12288,128,0.006689777804745569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,8192,0.03786844346258376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,8192,0.020961778031455144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,8192,0.031082666582531396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,65536,0.2041004498799642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,7168,0.019189332922299702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,7168,0.02735111117362976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,51200,0.09826311137941147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,6144,0.017494221528371174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,12288,0.04253155655331082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,5120,0.024661334024535284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,6144,0.023843554986847773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,4096,0.0207022229830424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,5120,0.015769778026474845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,5120,0.0206657780541314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,3584,0.017814222309324477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,4096,0.01384800010257297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,4096,0.01777866648303138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,3072,0.015824889143308003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,3584,0.013071111506885953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,7168,0.03313422203063965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,3584,0.01612444387541877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,2560,0.013765333427323235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,3072,0.01204533295498954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,3072,0.014808000789748298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,2048,0.01180355581972334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,2560,0.01073333372672399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,1536,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,2560,0.013272888958454132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,128,0.007320000065697565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,2048,0.008824889030721452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,1024,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,2048,0.011901333100265928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,1536,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,768,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,1024,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,1536,0.010274666878912184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,512,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,1024,0.008264000217119852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,256,0.004074666649103165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,768,0.007323555648326874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,128,0.003488000068399641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,512,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,768,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,10240,0.03646399908595615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,256,0.006694222076071634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,64,0.0034924443397257063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,512,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,32,0.0036968886852264404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,10240,128,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,256,0.0063279999627007385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,128,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,65536,0.222417778438992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,64,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,10240,32,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,51200,0.17086488670772979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,16384,0.05757244427998861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,65536,0.07497421900431316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,12288,0.04419822163052029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,65536,0.16116889317830405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,51200,0.12727555963728163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,16384,0.023740443918440077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,10240,0.0373324453830719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,16384,0.04451910985840691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,10240,6144,0.028813332319259644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,8192,0.031269331773122154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,7168,0.027452444036801655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,12288,0.03453600075509813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,10240,0.016861332787407767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,6144,0.023958222733603582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,8192,0.014892444014549255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,8192,0.025016888976097107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,7168,0.013861333330472311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,5120,0.020436444216304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,6144,0.012705778082211813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,7168,0.021566222111384075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,4096,0.017133333616786532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,6144,0.019648889700571697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,5120,0.011699555648697747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,3584,0.014929778046078153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,5120,0.017251556118329365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,4096,0.010763555765151978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,3072,0.013550221920013428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,3584,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,2560,0.011565333439244164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,3584,0.013723555538389417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,3072,0.008646222452322641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,2048,0.010016889207892949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,3072,0.012835555606418185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,2560,0.008292444050312042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,1536,0.008179555336634317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,2048,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,51200,0.060418665409088135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,2048,0.010412444670995077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,12288,0.019168888529141743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,1024,0.006492444624503453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,1536,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,768,0.005199111170238919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,1536,0.0087333329849773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,1024,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,10240,0.0295413335164388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,512,0.0038906666967603895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,1024,0.007326222128338284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,256,0.003474666840500302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,768,0.006517333288987477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,512,0.006246222390068903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,768,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,128,0.0028844444702068963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,512,0.006492444624503453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,256,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,64,0.0028479999552170434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,8192,32,0.0028524444335036804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,256,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,4096,0.015124445160230001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,128,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,64,0.005990222096443176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,65536,0.20408177375793457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,32,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,51200,0.16010133425394693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,8192,2560,0.011680888632933298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,65536,0.07186133331722684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,65536,0.15652088324228922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,12288,0.04248088929388258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,51200,0.12342222531636556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,16384,0.02264533274703556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,10240,0.03552088803715176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,16384,0.04341244366433886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,12288,0.01850755512714386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,8192,0.029907554388046265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,10240,0.01646577815214793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,12288,0.03350222110748291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,10240,0.028512001037597656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,7168,0.02625422179698944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,8192,0.014596444037225513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,8192,128,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,8192,0.023981332778930664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,6144,0.022611555125978258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,7168,0.013408000270525614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,5120,0.01906044450071123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,7168,0.02125688890616099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,4096,0.016192888220151264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,6144,0.0192586663696501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,5120,0.016882666283183627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,4096,0.010041777458455827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,3584,0.014127999544143677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,16384,0.054232001304626465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,4096,0.014836443795098199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,51200,0.05827111005783081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,3072,0.012837332983811697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,3584,0.009214222431182861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,3584,0.013755555782053204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,3072,0.008800889054934183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,3072,0.012657778130637275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,2048,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,1536,0.007958222594526079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,2560,0.008077333370844523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,2560,0.011437333292431302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,2048,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,1024,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,1536,0.0069368887278768755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,768,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,1536,0.008547555241319869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,1024,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,512,0.004062222109900581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,1024,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,768,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,256,0.003220444545149803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,768,0.006979555719428592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,512,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,5120,0.011394666300879585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,128,0.002814222127199173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,512,0.006346666564544042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,256,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,64,0.0031742221779293488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,32,0.0031893334041039148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,256,0.005983110931184556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,128,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,7168,2560,0.011009777585665384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,65536,0.15145777331458196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,64,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,32,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,51200,0.12265333864423965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,65536,0.06968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,65536,0.11910133891635472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,16384,0.04748533169428507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,51200,0.09398221969604492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,7168,2048,0.009850666754775578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,12288,0.03522488805982802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,16384,0.02183644473552704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,10240,0.029489778810077246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,12288,0.017940445078743827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,6144,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,12288,0.02610933283964793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,8192,0.024870221813519795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,10240,0.015947555502255756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,7168,0.021532444490326777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,10240,0.022641777992248535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,8192,0.014102222190962898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,6144,0.019015999303923715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,8192,0.01923733287387424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,7168,0.013066666821638743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,5120,0.01612711118327247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,6144,0.012120000190205045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,6144,0.015499555402331881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,4096,0.013648000028398303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,5120,0.010818666881985135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,5120,0.013774221969975365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,3584,0.011965333587593503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,4096,0.009499555660618676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,4096,0.01255555533700519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,51200,0.05636799997753567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,3072,0.010839111275143094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,3584,0.008808000220192803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,2560,0.009635555247465769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,16384,0.033879111210505165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,3072,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,2560,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,2560,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,7168,128,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,2048,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,7168,0.017106667160987854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,1536,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,2048,0.008408889174461365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,1024,0.005488889084921942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,1536,0.006986666884687211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,1024,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,1536,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,1024,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,512,0.0035333335399627686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,768,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,768,0.006301333506902059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,256,0.0028986665937635633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,512,0.00599644457300504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,3584,0.011459555890825061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,512,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,128,0.0025351111673646504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,3072,0.008191111187140146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,256,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,64,0.002472888885272874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,6144,128,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,256,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,2048,0.008068444828192392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,128,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,64,0.00564444437623024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,65536,0.1438551081551446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,6144,32,0.005879111174080107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,51200,0.1163440015580919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,65536,0.06687821944554646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,16384,0.043087111579047314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,65536,0.11569066842397054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,768,0.004436444491147995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,51200,0.05411555700831943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,12288,0.03306400113635593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,51200,0.09041155709160699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,10240,0.02812622321976556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,16384,0.020785777105225455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,12288,0.01721244388156467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,8192,0.023072888453801472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,16384,0.03271377748913235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,10240,0.01566755606068505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,7168,0.020270221763186984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,12288,0.0251528885629442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,10240,0.02160266704029507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,8192,0.013771555489963956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,7168,0.012728888955381183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,5120,0.015088889333936902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,8192,0.018966222802797954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,7168,0.016753777861595154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,4096,0.012805333567990197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,6144,0.01163733336660597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,6144,0.015509333875444202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,6144,32,0.0025964443468385274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,5120,0.010122666756312052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,3072,0.01036177741156684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,4096,0.00888355573018392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,5120,0.013785777820481194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,2560,0.008959111240175035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,4096,0.012445333103338877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,3072,0.008013333711359236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,3584,0.011374221907721626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,2048,0.007762666377756331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,3072,0.010486221975750394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,2560,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,1536,0.006433777925040986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,2560,0.009291555318567488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,2048,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,1024,0.00509333320789867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,2048,0.008261333737108443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,1536,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,768,0.004217777815130022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,1024,0.006568000134494569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,1536,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,512,0.0034844444857703317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,1024,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,768,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,768,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,6144,0.01755466726091173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,512,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,128,0.002543111021320025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,256,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,3584,0.01111288865407308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,64,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,128,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,32,0.0027662221756246355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,3584,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,65536,0.11084089014265273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,64,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,32,0.00565422202150027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,51200,0.08756888575024074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,65536,0.06470221943325467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,65536,0.10979111327065362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,5120,256,0.002954666606254048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,5120,512,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,51200,0.0517866677708096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,51200,0.08541866805818345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,12288,0.02773688899146186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,16384,0.020289777053727042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,16384,0.0306328899330563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,10240,0.02256533337963952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,256,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,12288,0.016764443781640794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,12288,0.02420711186197069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,8192,0.018844443890783522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,7168,0.016507556041081745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,5120,128,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,10240,0.015153777268197803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,6144,0.014359111587206522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,8192,0.013175999952687157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,7168,0.012254222399658628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,8192,0.018232888645595975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,6144,0.01107377807299296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,4096,0.011002666420406766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,6144,0.01514488955338796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,5120,0.009658666948477427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,3584,0.009639999932712978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,5120,0.013352000051074557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,4096,0.008636444807052612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,3072,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,4096,0.011890666352377998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,3584,0.00831200016869439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,16384,0.03418755531311035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,2560,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,3584,0.010760000182522668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,3072,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,2048,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,3072,0.009438222481144799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,2560,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,1536,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,2560,0.00867733359336853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,2048,0.007249777515729268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,1024,0.004176889028814104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,1536,0.006911110960774952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,2048,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,768,0.0034844444857703317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,1024,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,1536,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,512,0.0031537777847713898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,10240,0.021176889538764954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,256,0.002518222149875429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,5120,0.012498666842778524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,768,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,512,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,7168,0.016508445143699646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,128,0.0027182222240500977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,64,0.00249422217408816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,512,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,256,0.005873777800136142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,4096,32,0.002528888897763358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,128,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,256,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,65536,0.10273777776294285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,128,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,64,0.00553955551650789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,51200,0.08188444375991821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,32,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,16384,0.033011555671691895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,65536,0.06330933173497517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,65536,0.10610222154193455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,12288,0.026022222306993272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,51200,0.05056533217430115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,51200,0.08371555805206299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,10240,0.02129511038462321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,16384,0.03010222315788269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,12288,0.023617777559492324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,8192,0.017856889300876193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,10240,0.020621332857343886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,7168,0.01585155559910668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,4096,1024,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,8192,0.012921777864297232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,6144,0.013809777796268463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,8192,0.017879999346203275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,7168,0.011710222396585675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,7168,0.016152888536453247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,5120,0.012096888489193387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,6144,0.010484444598356882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,6144,0.014602666099866232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,4096,0.010610666539933948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,5120,0.009304888546466827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,5120,0.013182222015327878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,3584,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,4096,0.00868800034125646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,4096,0.011358222199810876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,16384,0.019938665959570143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,3072,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,3584,0.008294222255547842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,2560,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,3584,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,3072,0.00793511089351442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,12288,0.016501333978441026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,2048,0.006500444478458828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,3072,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,2560,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,1536,0.00535999983549118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,2560,0.008637333081828224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,2048,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,4096,768,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,1024,0.004190222256713443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,768,0.0037466668420367767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,2048,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,1536,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,512,0.003212444484233856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,1024,0.006454222318198945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,1536,0.006986666884687211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,256,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,768,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,1024,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,128,0.002461333241727617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,512,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,768,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,256,0.005662222289376789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,64,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3584,32,0.0022400000856982335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,256,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,65536,0.07717333237330119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,128,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,64,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3584,32,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,51200,0.06145333581500583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,16384,0.02479022244612376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,65536,0.062294224898020424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,65536,0.06812177764044867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,12288,0.02055377761522929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,51200,0.04973599976963467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,51200,0.05339644352595011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,10240,0.0148053334818946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,16384,0.02049422264099121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,12288,0.01649777756796943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,8192,0.016689777374267578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,12288,0.01681600014368693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,10240,0.014558222558763293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,8192,0.012077333198653327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,10240,0.014825777875052558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,8192,0.01293688929743237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,6144,0.012874667015340594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,7168,0.01108000013563368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,7168,0.011923555698659686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,6144,0.010114666488435533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,6144,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3584,128,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,4096,0.009911999934249455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,5120,0.009322666459613377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,5120,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,3584,0.008917333351241218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,4096,0.008424888882372115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,3072,0.007791111038790808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,4096,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,3584,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,2560,0.0069884442620807225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,3584,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,3072,0.00790666706032223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,2048,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,16384,0.019679110911157396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,3072,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,1536,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,2560,0.007589333587222629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,2560,0.006986666884687211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,2048,0.007239999870459239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,1024,0.0038577777643998465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,1536,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,2048,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,768,0.0034382223255104492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,1536,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,512,0.0028942223224374983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,1024,0.0064239998658498125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,768,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,5120,0.01146666705608368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,256,0.0026568888376156488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,768,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,512,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,128,0.002452444492114915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,512,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,64,0.002238222294383579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,32,0.0025271111064487034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,3072,128,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,128,0.005187555733654234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,64,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,65536,0.07321777608659533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,32,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,10240,0.018505778577592637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,51200,0.058670222759246826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,65536,0.06135644515355428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,65536,0.06525955597559611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,3072,7168,0.014479999740918478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,16384,0.023948444260491267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,12288,0.01995822290579478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,51200,0.048615998691982694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,51200,0.05115555392371284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,16384,0.01941688855489095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,10240,0.017086222767829895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,16384,0.019896000623703003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,8192,0.014174222946166992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,12288,0.016172443826993305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,10240,0.014334221680959066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,7168,0.012714666624863943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,8192,0.011754666765530905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,256,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,8192,0.012729778057999082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,6144,0.011315555208259158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,7168,0.010704889065689512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,5120,0.009887110855844285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,7168,0.0114702218108707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,6144,0.010040889183680216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,4096,0.008647999829716152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,6144,0.010605333579911126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,5120,0.009320889082219865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,3584,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,5120,0.009482666850090027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,4096,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,3072,0.007172444628344641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,3584,0.008303110798199972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,3584,0.007889777421951294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,2560,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,3072,1024,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,3072,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,2048,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,10240,0.013795555465751223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,2560,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,1536,0.004535111288229625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,2560,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,2048,0.006986666884687211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,1024,0.003871110992299186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,2048,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,1536,0.00674311113026407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,768,0.0035235554807715942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,1536,0.006223111102978389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,512,0.0029155556112527847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,1024,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,768,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,256,0.0025235555238193935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,768,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,512,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,4096,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,512,0.0053093334039052325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,128,0.0026977778308921387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,256,0.005766222046481238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,64,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2560,32,0.0021564445147911706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,128,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,128,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,65536,0.057803557978736035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,12288,0.016061334146393668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,64,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,32,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,3072,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,51200,0.046747555335362755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,16384,0.01956622302532196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,65536,0.0595688886112637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,12288,0.015885333220163982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,51200,0.04730222291416592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,51200,0.04929955469237434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,16384,0.01901511185699039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,10240,0.015096889601813423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,16384,0.01939644416173299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,12288,0.015166223049163818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,8192,0.014176888598336114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2560,1024,0.0063075555695427795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,10240,0.013471110827393003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,12288,0.0159288893143336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,7168,0.014140443669425117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,10240,0.014003554979960123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,6144,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,7168,0.010741333166758219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,8192,0.012210666305489011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,5120,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,7168,0.011047110789351992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,6144,0.010051555931568146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,4096,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,6144,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,5120,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,3584,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,5120,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,4096,0.008515555825498369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,3072,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,4096,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,3584,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,2560,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,3072,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,3072,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,2048,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,65536,0.06245599852667915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,2560,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,1536,0.004239999999602635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,2560,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,2048,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,1024,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,2048,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,1536,0.006665777828958299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,768,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,8192,0.011392888923486074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,1536,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,512,0.0028231110837724474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,1024,0.005724444571468566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,768,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,256,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,768,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2560,256,0.005193777796294955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,512,0.005342222336265776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,256,0.005712000032265981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,3584,0.008259555531872643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,64,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,128,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,32,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,64,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,32,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,65536,0.04435999857054817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,65536,0.058118224143981934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,1024,0.006310222049554189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,65536,0.0596675541665819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,51200,0.04657599992222256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,2048,512,0.005882666756709416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,2048,128,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,16384,0.015941333439615037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,51200,0.04753955536418491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,12288,0.013038222160604266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,256,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,16384,0.018402665853500366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,16384,0.01886666648917728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,2048,128,0.004962666581074397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,8192,0.012173333101802401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,12288,0.015671110815472074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,7168,0.011423110961914062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,10240,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,10240,0.013506666653686099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,6144,0.010334221853150262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,7168,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,8192,0.011599111060301462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,7168,0.010409778190983666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,6144,0.009840000006887648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,6144,0.009639110830095079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,4096,0.0075937774446275495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,51200,0.03577244281768799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,5120,0.009122666385438707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,5120,0.008964444200197855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,3584,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,4096,0.008604444563388824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,3072,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,4096,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,3584,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,2560,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,3584,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,3072,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,2048,0.00453955555955569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,10240,0.011693333586057028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,2560,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,1536,0.003903999924659729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,2560,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,2048,0.007100444700982835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,1024,0.0031822222388452957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,8192,0.01124533348613315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,1536,0.00666933341158761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,768,0.0029128889242808023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,1536,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,1024,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,5120,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,512,0.00253866674999396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,1024,0.005881777654091517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,768,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,256,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,768,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,512,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,256,0.005719999886221356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,512,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,256,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,32,0.0021991110924217436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,128,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,128,0.004951111144489712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,65536,0.032285332679748535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,64,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,32,0.005200888961553574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,51200,0.026716444227430556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,65536,0.0572648909356859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1536,12288,0.014167111780908374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,16384,0.013752000199423896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,65536,0.057271997133890785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,51200,0.045885334412256874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,2048,0.0064604443808396654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,12288,0.012206222448084088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,51200,0.04513955447408888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,16384,0.016767111089494493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,10240,0.011026666396194033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,16384,0.01852888862291972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,8192,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,12288,0.014101333088344999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,10240,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,12288,0.014472888575659858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,7168,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,8192,0.011254222028785281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1536,128,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,6144,0.008923555413881937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,8192,0.011023110813564725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,7168,0.010496888723638324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,5120,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,7168,0.010321777727868822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,4096,0.0069919998447100324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,5120,0.009328888522254096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,6144,0.009407999614874521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,3584,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,5120,0.008651555412345463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,4096,0.008504889077610439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,3072,0.005872888697518243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,4096,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1536,3072,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,3584,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,2560,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,3584,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,2048,0.004559111264016894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,2560,0.0074648890230390764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,3072,0.007270221908887227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,1536,0.0038933331767717996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,2560,0.00674933319290479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,2048,0.007224889265166388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,2048,0.0064604443808396654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,1024,0.0034755555291970572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,1536,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,768,0.0031822222388452957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,1536,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,10240,0.012717333104875354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,1024,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,512,0.0028293333533737394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,1024,0.0057297779454125305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,256,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,768,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,512,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,128,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,512,0.0052933332820733385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,256,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,64,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,256,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,128,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,1024,32,0.0022088889446523455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,128,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,65536,0.029071112473805744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,51200,0.023755555351575215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,64,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,1024,32,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,65536,0.05709422296947903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,16384,0.011732444167137146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,65536,0.055994669596354164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,51200,0.045591112640168935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,12288,0.010442666709423065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,16384,0.016732444365819294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,10240,0.011571555501884885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,16384,0.017716444200939603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,12288,0.014074666632546319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,8192,0.01036888857682546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,12288,0.014167999227841696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,768,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,10240,0.012759110993809171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,7168,0.009633777870072259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,8192,0.0110871113008923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,6144,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,6144,0.008800889054934183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,8192,0.010839111275143094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,7168,0.010668444136778513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,5120,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,6144,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,7168,0.010231999887360467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,4096,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,6144,0.009571555587980483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,3584,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,5120,0.00867022242810991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,4096,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,3072,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,1024,3072,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,4096,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,3584,0.008216000265545314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,2560,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,3584,0.007625777688291337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,3072,0.007711110843552484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,2048,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,51200,0.04450311263402303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,3072,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,1536,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,2048,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,2560,0.0069004446268081665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,1024,0.003576888806290097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,2048,0.006450666735569636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,1536,0.006588444527652528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,768,0.003177777760558658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,10240,0.012747555143303342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,1536,0.005976000179847081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,1024,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,512,0.0026106666773557663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,1024,0.005777777896987067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,256,0.0021946666141351065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,768,0.006138666636413998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,512,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,128,0.0021955555097924336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,512,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,256,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,5120,0.009007111191749573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,256,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,128,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,32,0.0019448887970712448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,128,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,65536,0.023543111152119104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,64,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,51200,0.018569777409235638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,32,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,65536,0.05671199825074938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,16384,0.012159110771285163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,768,2560,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,65536,0.054699556695090405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,12288,0.010429333481523724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,51200,0.04383111000061035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,16384,0.016799110505315993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,51200,0.04390133420626322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,10240,0.009560888840092553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,16384,0.01719288859102461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,8192,0.008443555898136562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,12288,0.014044443766276041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,10240,0.01276622215906779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,7168,0.007912000020345053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,10240,0.012485333614879184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,8192,0.011072888970375061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,768,768,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,6144,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,8192,0.010841777755154504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,5120,0.007325333025720384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,7168,0.010473777850468954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,768,64,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,7168,0.01036088913679123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,4096,0.006725333217117522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,5120,0.009000000026490953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,6144,0.009369778136412302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,3584,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,5120,0.008678221868144141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,3072,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,4096,0.00793511089351442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,3584,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,2560,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,3584,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,3072,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,3072,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,2048,0.004232888834344016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,2560,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,1536,0.003843555433882607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,2560,0.006784889019197888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,12288,0.01385600037044949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,2048,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,1024,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,768,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,1536,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,1024,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,1536,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,512,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,1024,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,256,0.0021608887861172357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,768,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,512,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,128,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,6144,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,256,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,64,0.0018639999131361644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,256,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,512,32,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,4096,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,65536,0.019476445184813607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,32,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,51200,0.016529776983790927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,65536,0.05395644572046068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,51200,0.039741333987977766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,2048,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,16384,0.01107377807299296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,768,0.005978666659858491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,16384,0.016688000824716356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,12288,0.009138666921191746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,16384,0.016923555069499545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,10240,0.008329778081840938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,512,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,12288,0.013924444715181986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,12288,0.014058666096793281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,512,128,0.005632888939645555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,512,128,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,10240,0.01258044441541036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,7168,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,8192,0.011145778000354767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,8192,0.0107342220014996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,6144,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,65536,0.04966488811704847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,7168,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,5120,0.008024888734022776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,6144,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,51200,0.04314133193757799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,6144,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,4096,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,5120,0.009020444419648912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,5120,0.008640888664457533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,3584,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,4096,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,3072,0.005542222410440445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,4096,0.007795555724038019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,3584,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,2560,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,3584,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,3072,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,8192,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,10240,0.012574222352769641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,2560,0.007381333245171442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,2560,0.006725333217117522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,1536,0.0035173334181308746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,2048,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,1024,0.003169777699642711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,1536,0.006656888872385025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,768,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,1536,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,7168,0.010055999788973067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,1024,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,512,0.002491555487116178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,1024,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,768,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,768,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,512,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,128,0.0018915554715527429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,512,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,256,0.005724444571468566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,64,0.001806222316291597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,256,0.004961777892377642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,256,128,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,65536,0.023576888773176406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,128,0.004955555415815777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,64,0.004947555561860402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,32,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,65536,0.049307554960250854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,51200,0.020615110794703167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,2048,0.004243555582231945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,16384,0.013759111364682516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,65536,0.05116533239682516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,12288,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,51200,0.03974044322967529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,16384,0.016555555992656283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,51200,0.04107022285461426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,10240,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,2048,0.0063893331421746155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,12288,0.014091556270917257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,8192,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,12288,0.013791999883121915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,10240,0.012793777717484368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,8192,0.011094222466150919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,10240,0.012390221986505719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,8192,0.010812444819344414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,6144,0.007256000406212277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,7168,0.010471111370457543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,7168,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,5120,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,6144,0.009642666412724389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,256,32,0.0018133332745896445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,4096,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,6144,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,5120,0.009045333498054082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,3584,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,4096,0.008284444610277811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,5120,0.008642666869693333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,4096,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,256,3072,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,2560,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,3584,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,3072,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,2048,0.0041831110914548235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,2560,0.007479111353556315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,16384,0.016527111331621807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,2560,0.006559111177921295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,1536,0.003547555456558863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,2048,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,7168,0.00666844430896971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,1024,0.0028533333291610083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,1536,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,768,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,1536,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,1024,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,1024,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,768,0.0052986666560173035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,512,0.0025315555847353404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,768,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,3072,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,512,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,256,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,3584,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,512,0.00518488883972168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,256,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,3072,0.007105777661005656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,256,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,64,0.0018079999006456798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,161,128,128,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,32,0.0018159999615616268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,128,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,65536,0.016153777639071148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,51200,0.013778666655222574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,2048,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,64,0.0047626665068997275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,16384,0.006984000404675801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,128,32,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,12288,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,10240,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,65536,0.05080355538262261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,8192,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,16384,0.01651733285850949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,12288,0.013574221895800697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,10240,0.012108444339699216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,6144,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,8192,0.010440888504187265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,5120,0.006495999793211619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,7168,0.009938666390048133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,128,128,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,6144,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,3584,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,5120,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,3072,0.0053004444473319584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,4096,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,2560,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,3584,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,2048,0.004214222232500712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,1536,0.003580444388919406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,3072,0.006895111252864202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,1024,0.0029253332565228143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,2560,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,51200,0.04113066527578566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,2048,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,512,0.00254488881263468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,1024,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,7168,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,256,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,128,0.0018702221827374564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,768,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,64,0.0018115554832749898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,512,0.005132444616821078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,32,0.0018079999006456798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,256,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,4096,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,128,0.004655999855862724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,51200,0.013441777891582914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,64,0.005064888960785336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,16384,0.011340444286664328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,12288,0.009019555317031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,32,0.004701333327425851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,10240,0.006346666564544042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,65536,0.0506933331489563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,8192,0.005628444254398346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,16384,0.016837333639462788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,12288,0.013426666458447775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,7168,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,6144,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,10240,0.012110222544935016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,5120,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,64,768,0.002790222151411904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,4096,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,64,1536,0.005976888868543837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,7168,0.009728888670603434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,3584,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,6144,0.00925511121749878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,3072,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,5120,0.008303110798199972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,2560,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,2048,0.0042239998777707415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,4096,0.007690666450394525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,1536,0.0035013332962989807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,3584,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,1024,0.0028693332440323303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,3072,0.007280888656775157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,768,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,2560,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,65536,0.015951999359660678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,512,0.0024480000138282776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,256,0.0021359999146726397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,1536,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,128,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,768,0.00553688903649648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,512,0.005104889058404499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,51200,0.040764444404178195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,256,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,32,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,128,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,64,0.004599110947714912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,32,0.004624888714816836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,2048,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,1024,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,161,32,64,0.0018088887963030075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,51200,1.0884346432156033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,51200,0.8570346832275391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,16384,0.3486559920840793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,161,32,8192,0.010457778142558204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,16384,0.2698391013675266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,12288,0.2667164537641737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,16384,0.13623555501302084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,12288,0.20218843883938262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,10240,0.23311734199523926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,12288,0.10429155826568604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,10240,0.1770311064190335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,51200,0.40102577209472656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,8192,0.19407555792066786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,10240,0.08714489142100017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,8192,0.13957422309451634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,7168,0.16698488924238417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,8192,0.07181866963704427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,7168,0.11997777885860866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,12288,0.2267306645711263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,7168,0.06360444757673475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,10240,0.18848889403873018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,6144,0.14282755057017008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,16384,0.30181511243184406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,5120,0.12143466207716201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,6144,0.05448266532686022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,6144,0.10385777552922566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,7168,0.13721244864993626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,4096,0.10132177670796712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,5120,0.08706133233176337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,5120,0.04635999931229485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,3584,0.08673244714736938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,4096,0.07334044244554308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,4096,0.039846221605936684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,5120,0.09968533118565877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,4096,0.08215555879804823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,3584,0.03519466519355774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,3584,0.06326666805479261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,3072,0.07580533292558458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,2560,0.06244977977540758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,3584,0.07338311274846394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,3072,0.03128000100453695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,8192,0.15004711680942112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,2048,0.05142133434613546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,3072,0.06532444556554158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,2560,0.027637332677841187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,2560,0.057413332992129855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,2048,0.04087466663784451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,2048,0.02383111086156633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,2560,0.04797777864668104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,1536,0.03940266701910231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,6144,0.12006489435831706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,1024,0.02849777870708042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,2048,0.04930577675501505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,1536,0.01889866590499878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,1536,0.032271110349231295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,768,0.021648888786633808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,1024,0.015607999430762397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,1024,0.02530844509601593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,512,0.015376889043384127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,1024,0.03423555692036947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,768,0.014078223043017916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,768,0.025511110822359722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,256,0.009732444253232744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,768,0.031882666879230075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,3072,0.055035554700427584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,512,0.01772799922360314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,128,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,512,0.013175999952687157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,512,0.0251555558707979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,256,0.011710222396585675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,256,0.015480000111791821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,51200,0.9397893481784396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,64,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,65536,32,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,128,0.011698666546079846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,128,0.015057777365048727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,128,0.016474665866957772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,64,0.014732443624072604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,65536,32,0.015216888652907478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,1536,0.0429440008269416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,51200,0.8974515067206489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,65536,0.8761528862847222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,65536,1.162642690870497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,51200,0.6664026578267416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,256,0.02077244387732612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,16384,0.28992001215616864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,65536,0.410883559121026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,12288,0.2214648856057061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,16384,0.10803644524680243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,16384,0.21880443890889487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,12288,0.20127377245161268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,12288,0.0815804468260871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,51200,0.31665155622694224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,10240,0.14795022540622288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,16384,0.2425004376305474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,10240,0.06911022133297391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,8192,0.1375839975145128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,12288,0.1855457756254408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,10240,0.1542444494035509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,8192,0.11274488766988118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,10240,0.16861332787407768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,7168,0.11998132864634196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,7168,0.05019111103481717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,6144,0.10568088955349392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,8192,0.12514221668243408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,6144,0.08345688713921441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,7168,0.11194399992624919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,5120,0.09219821956422593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,6144,0.09611022472381592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,8192,0.05685066514545017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,6144,0.04416355490684509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,4096,0.07582222090827094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,5120,0.07060177458657159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,5120,0.03793066740036011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,7168,0.09585066636403401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,4096,0.0581644442346361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,3584,0.06411200099521212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,5120,0.08208355638715956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,4096,0.03264000018437704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,3072,0.05693333016501533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,4096,0.0690222250090705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,3584,0.02930755416552226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,51200,0.7351013289557563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,2560,0.04811911119355095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,3072,0.04523288872506884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,3072,0.026222222381167944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,2048,0.04032889008522034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,3072,0.05683466460969713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,2560,0.02344177828894721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,2560,0.04243110948138767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,1536,0.03105866577890184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,2560,0.049866666396458946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,2048,0.033360888560613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,2048,0.019842666056421068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,1024,0.02331111166212294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,2048,0.043509334325790405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,1536,0.016350222958458793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,1536,0.027802666028340656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,65536,1.0196524726019967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,768,0.017661333084106445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,1536,0.03694577680693732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,1024,0.013771555489963956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,512,0.011928888658682505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,1024,0.030108445220523413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,768,0.026846221751636926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,768,0.026467555099063452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,3584,0.05125066637992859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,512,0.01588711142539978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,3584,0.06312533219655354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,512,0.022653333014912073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,128,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,512,0.013215111361609565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,256,0.01036444471942054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,128,0.010488000180986194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,128,0.012463111016485425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,64,0.005372444374693765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,32,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,128,0.015024888846609326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,64,0.01273155543539259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,1024,0.021007999777793884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,32,0.013215111361609565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,65536,0.37905510266621906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,51200,768,0.01277955538696713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,51200,256,0.008408889174461365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,65536,0.30305777655707467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,51200,0.29601067966885036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,65536,0.13976800441741943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,51200,256,0.013441777891582914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,16384,0.09735466374291314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,51200,0.23476089371575248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,51200,256,0.018915555543369718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,12288,0.07556088765462239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,16384,0.07946044206619263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,16384,0.0404426654179891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,12288,0.060264892048305936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,10240,0.06212266948488024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,16384,0.08567821979522705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,12288,0.06964088810814752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,10240,0.027666666441493567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,10240,0.051282667451434664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,8192,0.05307288964589437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,10240,0.05951555569966634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,8192,0.04376799861590067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,7168,0.0454346670044793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,8192,0.023582221733199224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,8192,0.05067288875579834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,65536,0.32239288753933376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,7168,0.020976889464590285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,6144,0.039072000318103366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,7168,0.03789955708715651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,51200,0.10980444484286839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,5120,0.033987555238935686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,7168,0.04559822214974297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,6144,0.018903111418088276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,6144,0.033534222178988986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,6144,0.041618665059407554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,4096,0.028605331977208454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,5120,0.016741333736313712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,5120,0.02903111113442315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,5120,0.03654222355948554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,3584,0.024989333417680528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,4096,0.014760888285107084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,4096,0.024259555670950148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,3072,0.021868444151348535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,4096,0.032103111346562706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,3584,0.013760888742076026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,3584,0.021927111678653296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,2560,0.018199111024538677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,3584,0.029373334513770208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,3072,0.012847999731699625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,3072,0.019494222270117868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,51200,0.24865600797865126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,2048,0.014927999840842353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,3072,0.026603556341595117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,2560,0.011791999969217511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,2560,0.01703999936580658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,1536,0.012033777932325998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,2560,0.02424888809521993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,2048,0.010654222634103564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,1024,0.009328000247478485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,2048,0.014890667464998035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,1536,0.009069333473841349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,1536,0.01274577776590983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,768,0.007715555528799693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,2048,0.021631111701329548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,1536,0.01923111081123352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,1024,0.008023999631404877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,1024,0.010770666930410596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,512,0.005991111199061076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,768,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,1024,0.016765332884258695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,768,0.015825778245925903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,256,0.0041884444653987885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,512,0.007229333122571309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,512,0.013443555268976422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,128,0.003535110917356279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,256,0.006868444383144379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,256,0.007333333293596904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,256,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,64,0.003170666595300039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,128,0.0069146665434042616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,16384,32,0.0032231110251612137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,128,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,128,0.00963644435008367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,64,0.007270221908887227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,32,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,12288,0.03182222114668952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,65536,0.273909330368042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,65536,0.217921773592631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,768,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,51200,0.2144897778828939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,16384,512,0.008092444803979661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,16384,0.0736702216996087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,51200,0.17073244518703887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,12288,0.055580443806118436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,16384,0.037681778271993004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,51200,0.10055910878711277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,16384,0.05788533555136787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,10240,0.046315554115507335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,12288,0.04452711012628344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,16384,0.08526311318079631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,12288,0.029712888929578993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,8192,0.03944799966282315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,12288,0.06719555457433064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,10240,0.025552888711293537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,10240,0.03815644317203098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,7168,0.033972445461485125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,10240,0.05881955226262411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,8192,0.021673777037196692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,8192,0.032680888970692955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,51200,0.23968532350328234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,6144,0.029595557186338637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,65536,0.3055182298024495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,8192,0.049787554475996226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,7168,0.019704000817404855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,5120,0.025407110651334126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,6144,0.017839999662505258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,7168,0.04482044445143806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,6144,0.02518400053183238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,4096,0.021746666895018682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,6144,0.040511998865339495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,5120,0.02163911031352149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,5120,0.01610133383009169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,3584,0.01867999964290195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,5120,0.03596977723969354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,4096,0.014472888575659858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,4096,0.01847555571132236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,3584,0.01331822243001726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,4096,0.03090311090151469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,3584,0.016763556334707472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,2560,0.013724444641007317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,3584,0.027963555521435205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,3072,0.012416000167528788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,3072,0.025968889395395916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,2560,0.013759111364682516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,2048,0.011911999848153857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,2560,0.011219555305110084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,7168,0.02863111098607381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,65536,0.12639199362860784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,2048,0.009763555394278632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,2048,0.012439111040698158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,1536,0.009928888744778102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,2048,0.02130222154988183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,1024,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,1536,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,1536,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,3072,0.016505777835845947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,768,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,1536,0.01904977858066559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,1024,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,1024,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,512,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,1024,0.016146666473812528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,3072,0.015126221709781222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,768,0.008616000413894653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,256,0.003777777983082665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,768,0.015170666906568738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,512,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,512,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,128,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,2560,0.023499555057949487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,256,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,256,0.006680000159475539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,64,0.003181333343187968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,256,0.011360888679822287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,12288,32,0.0033440000067154565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,128,0.006650666809744305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,128,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,64,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,65536,0.2447662353515625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,12288,32,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,51200,0.19245421886444092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,65536,0.20387466748555502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,65536,0.12330399619208442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,768,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,51200,0.15907022688123915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,16384,0.06579822301864624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,51200,0.09836710823906793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,512,0.013016000390052795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,16384,0.05510400070084465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,12288,0.05120355553097195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,16384,0.0851822230550978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,128,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,12288,0.028528001573350694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,10240,0.04169688953293694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,12288,0.04247733288341098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,8192,0.03495111068089803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,12288,0.06680888599819608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,65536,0.30321423212687176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,10240,0.02459999918937683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,10240,0.03656889001528422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,7168,0.03064889046880934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,51200,0.23708089192708334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,8192,0.020936000678274367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,8192,0.031140443351533677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,6144,0.026919111609458923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,8192,0.049215111467573375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,7168,0.027448889282014634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,5120,0.02316711180739933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,7168,0.044255110952589244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,6144,0.017468444175190397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,6144,0.023992000354660883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,4096,0.01974933346112569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,6144,0.039213332864973284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,16384,0.0359280010064443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,5120,0.015662221444977652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,5120,0.020663999848895602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,3584,0.016914667354689706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,4096,0.014075555735164218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,5120,0.03450489044189453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,4096,0.017616000440385606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,3072,0.0151555554734336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,4096,0.029767112599478826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,3584,0.016336888074874878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,3584,0.01309511148267322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,2560,0.012805333567990197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,3584,0.027452444036801655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,3072,0.011891555454995898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,3072,0.014868444866604276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,2048,0.011023110813564725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,2560,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,3072,0.025116443634033203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,10240,0.057849778069390186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,1536,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,2048,0.0088764445649253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,7168,0.019162666466501024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,2048,0.01186044431395001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,2048,0.020762667059898376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,1536,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,1536,0.010325333310498131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,768,0.006008000009589725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,1536,0.018443556295500863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,1024,0.007589333587222629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,1024,0.008053333395057255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,1024,0.014747555057207743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,512,0.005000000198682149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,768,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,768,0.014152889450391134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,512,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,512,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,512,0.012431999875439538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,2560,0.013382222089502545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,256,0.006567111031876669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,128,0.0031928889867332247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,2560,0.0229404436217414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,256,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,256,0.01128355579243766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,1024,0.007601777712504069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,64,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,32,0.0031555555760860443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,128,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,128,0.0063324446479479475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,128,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,64,0.006649777707126405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,65536,0.18191556135813394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,32,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,10240,768,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,51200,0.141720003551907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,65536,0.1622942156261868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,16384,0.04959466722276476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,65536,0.07479466994603474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,10240,256,0.003751111113362842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,51200,0.12670844131045872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,12288,0.0372977786593967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,51200,0.06034488810433281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,16384,0.023860444625218708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,16384,0.04462577899297079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,10240,0.03222044308980306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,16384,0.05104000038570828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,12288,0.019333332777023315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,12288,0.03434222274356418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,51200,0.1346551047431098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,65536,0.1678631173239814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,8192,0.02698844505680932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,12288,0.041393776734670006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,10240,0.029543999168607924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,7168,0.023406222462654114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,10240,0.016859556237856548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,10240,0.03654399845335219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,8192,0.015127110812399121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,7168,0.013736888766288757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,7168,0.021692444880803425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,8192,0.03149244520399306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,5120,0.01741155485312144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,7168,0.028592000404993694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,6144,0.019376888871192932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,6144,0.012703999876976013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,6144,0.02648266653219859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,4096,0.014922666880819531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,5120,0.017296888762050204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,5120,0.011514667007658215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,3584,0.012912888493802814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,4096,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,5120,0.023563555545277063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,4096,0.015159999330838522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,3072,0.011400000088744693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,3584,0.009694221946928237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,4096,0.021430222524536982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,2560,0.009672000176376766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,3584,0.013778666655222574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,3584,0.01980533368057675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,3072,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,3072,0.012842666771676807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,2048,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,3072,0.01868888901339637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,2560,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,1536,0.007086222370465596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,2560,0.011687999798191918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,2560,0.017286222842004564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,6144,0.020255999432669747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,2048,0.007705777883529663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,1024,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,8192,0.024923556380801733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,2048,0.016176889340082806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,1536,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,768,0.004552888787455029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,1536,0.013644444445768992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,1024,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,1024,0.007346666521496243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,512,0.003636444608370463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,1024,0.012078222301271228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,768,0.0063573333124319715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,768,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,256,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,768,0.012151999606026543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,512,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,512,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,128,0.0025537777692079544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,256,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,256,0.006116444451941385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,512,0.011142222417725457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,64,0.002471111093958219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,8192,32,0.0024986666523747975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,128,0.005958222266700532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,128,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,128,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,64,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,65536,0.1672595606909858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,32,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,2048,0.010423999693658618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,51200,0.12946044074164495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,8192,1536,0.009288888838556077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,65536,0.15629955132802328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,16384,0.04487199915779961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,65536,0.07157066795561048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,51200,0.12349066469404434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,51200,0.05823555257585314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,12288,0.03538399934768677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,16384,0.04306844539112515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,16384,0.022229333718617756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,16384,0.04974133438534207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,65536,0.15957778029971653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,256,0.010344888601038191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,10240,0.02958044409751892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,12288,0.018532445033391316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,51200,0.12544355127546522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,12288,0.040221333503723145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,10240,0.01643111142847273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,10240,0.028307557106018066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,7168,0.021835555632909138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,10240,0.03562044435077243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,8192,0.02416444487041897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,8192,0.030797332525253296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,7168,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,6144,0.018903111418088276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,7168,0.021126222279336717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,7168,0.027936001618703205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,5120,0.016149333781666227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,6144,0.012442666623327466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,6144,0.01920977731545766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,6144,0.02515022291077508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,4096,0.013815110756291283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,5120,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,12288,0.033647110064824425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,5120,0.023353776997990076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,8192,0.025285333395004272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,3584,0.01202755586968528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,4096,0.01479022204875946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,4096,0.010199111368921068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,4096,0.020512888828913372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,3072,0.01109155515829722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,3584,0.009085333181752099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,8192,0.01457244488928053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,3584,0.013743110828929476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,3072,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,3584,0.01960533360640208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,3072,0.012794666820102267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,2048,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,2560,0.00758844448460473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,3072,0.01837688850031959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,2560,0.011456888582971362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,1536,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,2560,0.016952888833151925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,2048,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,2048,0.009929777847396003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,1024,0.005379555539952383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,2048,0.015592000550693937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,1536,0.006979555719428592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,1536,0.008437333007653555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,768,0.004942222187916438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,1536,0.014198222094111972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,5120,0.01686755485004849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,1024,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,512,0.0035413333939181436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,1024,0.01186222251918581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,768,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,768,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,256,0.0029893333299292457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,768,0.011695110963450538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,512,0.00654666663871871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,128,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,256,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,2560,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,256,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,256,0.00998933364947637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,64,0.00249422217408816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,7168,32,0.002564444517095884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,128,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,128,0.005958222266700532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,128,0.008001777860853408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,64,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,32,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,65536,0.1178942256503635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,65536,0.06954933537377252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,7168,1024,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,512,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,51200,0.09237066904703777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,51200,0.05643466446134779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,512,0.011008000208271874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,65536,0.1488000022040473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,12288,0.032496889432271324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,16384,0.021760889225535925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,16384,0.03349422083960639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,51200,0.11842311753167047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,16384,0.047064890464146934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,10240,0.027377777629428442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,12288,0.02586311101913452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,65536,0.14920088979932997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,12288,0.038145777251985334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,8192,0.023388443721665278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,51200,0.11949333879682754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,10240,0.022258667482270136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,10240,0.03449689017401801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,7168,0.020392000675201416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,8192,0.014139556222491793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,8192,0.018944889307022095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,8192,0.029537777105967205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,6144,0.017684444785118103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,7168,0.013099555340078143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,16384,0.04523466693030464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,7168,0.016951999730534024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,5120,0.015160888433456421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,7168,0.02674399978584713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,6144,0.012083555261294046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,6144,0.015604444675975375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,4096,0.013374221821626028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,6144,0.024244444237815008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,5120,0.010999110837777456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,5120,0.014103111293580798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,3584,0.01145066652033064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,12288,0.017782222893502977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,5120,0.021918222308158875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,4096,0.009370666411187913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,3072,0.01053600013256073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,4096,0.012561777399645912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,3584,0.008943999807039896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,10240,0.01607288916905721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,4096,0.01977511081430647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,2560,0.009778666827413771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,3072,0.008200888832410177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,3072,0.010843555960390302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,2048,0.008408000071843466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,3072,0.01769244505299462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,2560,0.009654222263230218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,2560,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,1536,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,2560,0.016173332929611206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,2048,0.007253333098358578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,2048,0.008569777839713627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,1024,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,2048,0.014870221416155497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,1536,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,1536,0.012872888810104795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,768,0.004231111043029361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,1024,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,1024,0.011511111425028907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,512,0.004110222061475118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,768,0.006446222050322427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,768,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,768,0.011064888702498542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,3584,0.011535999675591787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,256,0.003212444484233856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,512,0.006100444330109491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,512,0.00601333338353369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,128,0.0028604444944196274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,512,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,256,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,256,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,64,0.0025271111064487034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,256,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,6144,32,0.0026231110095977783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,128,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,128,0.007672888537247975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,1536,0.0069146665434042616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,65536,0.13926843802134195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,64,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,32,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,1024,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,51200,0.12235466639200847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,65536,0.1166871123843723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,65536,0.06711822085910372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,16384,0.044772442844178945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,51200,0.09099911318884955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,51200,0.0539591113726298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,12288,0.03495999839570787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,3584,0.01874755488501655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,16384,0.020836444364653692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,16384,0.032834665642844305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,65536,0.14601955148908827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,51200,0.11645955509609646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,16384,0.046205331881841026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,12288,0.017363554901546903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,12288,0.025076443950335186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,8192,0.02457511093881395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,10240,0.015474667151769003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,10240,0.021543999512990315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,6144,128,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,12288,0.03722933265897963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,7168,0.021498666869269475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,10240,0.0324524442354838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,8192,0.013724444641007317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,6144,0.018779555956522625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,7168,0.012763555679056378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,8192,0.02851822310023838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,7168,0.017082666357358296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,5120,0.015812445018026564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,7168,0.025891555680169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,6144,0.015489778584904142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,6144,0.01143200033240848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,4096,0.013426666458447775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,5120,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,5120,0.013653332988421122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,5120,0.02146400014559428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,4096,0.012425777812798819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,3584,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,4096,0.009175111022260454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,10240,0.029489778810077246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,4096,0.019328888919618394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,3072,0.010621333287821876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,3584,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,3584,0.011396444506115384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,2560,0.009327111144860586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,3584,0.01814844376511044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,3072,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,3072,0.010788444843557147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,2048,0.007912889122962952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,3072,0.017342221405771043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,2560,0.007683555285135905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,2560,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,2560,0.015970667203267414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,1536,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,2048,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,2048,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,1024,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,6144,0.023733332753181458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,1536,0.006991110742092133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,1536,0.0074879998962084455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,768,0.0041804446114434134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,1536,0.012439111040698158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,1024,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,1024,0.00664444433318244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,512,0.003549333247873518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,1024,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,768,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,768,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,256,0.003176888864901331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,768,0.010755555497275459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,512,0.005995555470387141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,512,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,128,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,512,0.010063110954231685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,256,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,256,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,8192,0.01868266695075565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,64,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,128,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,5120,32,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,128,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,128,0.00758133331934611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,65536,0.10066399971644084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,64,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,2048,0.014224888549910652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,5120,32,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,65536,0.10927910937203301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,65536,0.06462222337722778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,16384,0.032219555642869734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,51200,0.051687111457188926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,51200,0.08613422181871201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,12288,0.025740444660186768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,65536,0.14552888605329725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,16384,0.020287111401557922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,16384,0.03131555517514547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,10240,0.019131556153297424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,16384,0.04571466644605001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,12288,0.024100444383091394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,12288,0.01682133310370975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,8192,0.016542222764756944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,12288,0.036706666151682533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,10240,0.015080889066060385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,10240,0.021016889148288306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,7168,0.014096000128322177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,10240,0.03200444579124451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,8192,0.01819644371668498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,8192,0.02794755498568217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,51200,0.07942755354775323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,6144,0.012395555774370829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,7168,0.025626666016048853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,6144,0.011020444333553314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,6144,0.01496355566713545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,6144,0.023464888334274292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,5120,0.011008888483047485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,51200,0.1156453291575114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,4096,0.010935111178292168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,5120,0.013424889081054263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,5120,0.021499555971887376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,256,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,4096,0.01205599970287747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,3584,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,4096,0.01891999940077464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,8192,0.01314133322901196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,3072,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,3584,0.00831288927131229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,7168,0.016520000166363187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,3584,0.01791200041770935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,3072,0.009881777895821465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,2560,0.006814222368929122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,3072,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,3072,0.016581333345837064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,2560,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,5120,0.009602666729026372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,2048,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,2560,0.014868444866604276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,2048,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,2048,0.00792266676823298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,4096,0.008772444393899705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,2048,0.013457777599493662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,1536,0.006685333533419504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,1024,0.0035422220826148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,7168,0.011944000091817645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,1536,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,1024,0.006532444308201472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,768,0.003175111073586676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,1024,0.011085333095656501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,2560,0.0075911109646161394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,768,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,512,0.0028320000403457214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,768,0.006147555592987273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,1536,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,512,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,256,0.002472888885272874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,512,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,512,0.010052444206343757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,1536,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,3584,0.010675555302037133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,256,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,256,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,256,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,64,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,128,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,32,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,128,0.005502222312821283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,128,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,64,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,4096,32,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,51200,0.07260000043445163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,65536,0.10635022322336833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,768,0.010417777631017896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,65536,0.06342044141557482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,16384,0.029410666889614526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,51200,0.05059822069274055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,65536,0.14470222261216906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,51200,0.08347911304897732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,12288,0.022966222630606756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,51200,0.1157991091410319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,16384,0.030341333813137476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,4096,128,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,10240,0.019928000039524503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,16384,0.0451128880182902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,12288,0.023669333921538457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,12288,0.016570667425791424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,1024,0.006451555424266391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,8192,0.015127999915017022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,10240,0.020639111598332722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,12288,0.03618311219745212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,65536,0.09223110808266534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,7168,0.013519110778967539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,10240,0.031409776873058744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,8192,0.012936000194814471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,8192,0.01752622259987725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,6144,0.0120319997270902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,8192,0.02762577765517765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,7168,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,7168,0.016023111012246873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,5120,0.010567111273606619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,6144,0.010427555276287926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,7168,0.02565244502491421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,6144,0.014811555544535318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,4096,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,6144,0.023359111613697473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,5120,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,5120,0.013117333253224691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,5120,0.020956445071432326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,4096,0.008766222331258986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,4096,0.011515555282433828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,16384,0.019941333267423842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,3072,0.00741511086622874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,4096,0.019267555740144517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,3584,0.008272888759771982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,3584,0.010084444450007545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,2560,0.006565333240562015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,3072,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,3072,0.009323555562231276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,2048,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,10240,0.014864888456132678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,3072,0.016157334049542744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,2560,0.00849511143234041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,2560,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,1536,0.004856889032655292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,2560,0.014771555860837301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,2048,0.0074471111098925276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,1024,0.003563555578390757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,2048,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,1536,0.006907555378145642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,2048,0.013763555222087435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,1536,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,768,0.0031822222388452957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,1536,0.01197866681549284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,1024,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,1024,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,512,0.0029226665695508323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,1024,0.011064888702498542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,3584,0.009740444521109263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,768,0.006120000034570694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,768,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,256,0.0025315555847353404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,512,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,512,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,128,0.002234666711754269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,512,0.010037333601050908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,256,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,256,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,64,0.0022115556316243275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,256,0.009304888546466827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3584,32,0.0021626665774318907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,128,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,128,0.005408888889683618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,65536,0.0769688884417216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,128,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,64,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3584,32,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,51200,0.06173155705134074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,65536,0.06767200099097358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,65536,0.0624106658829583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,16384,0.025251555773946974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,51200,0.05322844452328152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,51200,0.049842665592829384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,65536,0.14474666118621826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,12288,0.021290666527218286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,16384,0.020622221959961783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,51200,0.11544888549380833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,16384,0.019744889603720773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,768,0.010656889114114972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,10240,0.017829333742459614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,16384,0.04457777738571167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,12288,0.01686755485004849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,8192,0.017785777648289997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,10240,0.014821334017647637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,3584,0.01769333415561252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,12288,0.035967111587524414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,10240,0.014814222852389017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,7168,0.01532444523440467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,10240,0.031339555978775024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,8192,0.012274666792816587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,6144,0.01408266690042284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,8192,0.012873777912722694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,7168,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,8192,0.02737866673204634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,5120,0.012120889292822944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,7168,0.011962666279739805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,6144,0.0103502223889033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,7168,0.025439111722840205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,4096,0.00980799976322386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,6144,0.011071110765139261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,5120,0.009194666312800514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,6144,0.02292266653643714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,3584,0.008989333278603023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,5120,0.010054222411579555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,4096,0.00870666652917862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,5120,0.020821332931518555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,3072,0.008030222521887885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,4096,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,4096,0.018927110566033255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,2560,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,3584,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,3072,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,3584,0.016862221890025668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,3072,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,2048,0.005981333139869902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,3072,0.016146666473812528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,12288,0.016523554921150208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,2560,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,2048,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,2560,0.014470222923490735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,2048,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,1024,0.0038533334930737815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,2048,0.013495999905798169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,1536,0.006908444480763541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,1536,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,768,0.0033662221911880705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,1536,0.012667555775907306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,1024,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,1024,0.01105955574247572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,512,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,768,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,768,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,768,0.011036444041464063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,512,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,256,0.0026986667265494666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,512,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,512,0.010392888552612728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,128,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,256,0.00562755556570159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,256,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,2560,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,1536,0.004603555632962121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,256,0.009731555150614845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,64,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,3072,32,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,128,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,128,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,65536,0.06759733623928495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,64,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,32,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,1024,0.006517333288987477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,51200,0.05384799838066101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,65536,0.06519110997517903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,65536,0.0614008903503418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,16384,0.021946666969193354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,65536,0.14447644021775988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,51200,0.05096800128618876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,51200,0.0483822226524353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,3584,0.008007110820876228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,51200,0.11481332778930664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,16384,0.02012888921631707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,16384,0.019539554913838703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,10240,0.015860444969601102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,16384,0.044136888451046415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,12288,0.016114667057991028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,12288,0.016153777639071148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,8192,0.01646222174167633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,12288,0.03515555461247762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,3072,128,0.005030222237110138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,10240,0.014152000347773233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,10240,0.014352000421947904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,7168,0.014081777797804939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,10240,0.031311111317740545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,8192,0.0120328888297081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,8192,0.012650666965378655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,6144,0.012425777812798819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,7168,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,7168,0.011570666399266986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,8192,0.027128888501061335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,5120,0.01148000028398302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,7168,0.025008888708220586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,6144,0.0105262224872907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,4096,0.009761778016885122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,6144,0.023013333479563396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,5120,0.009372444616423713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,5120,0.009316444396972656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,3584,0.008600888980759515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,12288,0.018054222067197163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,4096,0.008622222476535374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,4096,0.008022222254011366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,3072,0.007687999970383114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,4096,0.01773066653145684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,3584,0.007888889147175683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,3584,0.01663466625743442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,2560,0.006828444285525217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,3072,0.007673777639865875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,3072,0.007385777930418651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,3072,0.015676443775494892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,2048,0.005492444253630108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,2560,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,2560,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,1536,0.004596444467703502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,2048,0.0069822221994400024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,2048,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,2048,0.013384888569513956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,6144,0.010436444646782346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,1024,0.0036799998746977914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,1536,0.006745777610275481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,1536,0.012103111379676394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,768,0.003170666595300039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,1024,0.006396444307433234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,1024,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,1024,0.01108088923825158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,512,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,768,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,768,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,768,0.010412444670995077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,256,0.0028568889117903183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,512,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,512,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,512,0.010462221999963125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,256,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,128,0.0021671110557185277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,256,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,2560,0.014099554883109199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,256,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,64,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2560,32,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,128,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,128,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,65536,0.051572445366117686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,1536,0.006325333482689328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,64,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2560,32,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,51200,0.04389511214362251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,65536,0.05937955776850382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,65536,0.061816891034444175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,16384,0.018034666776657104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,3584,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,51200,0.04913422134187487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,65536,0.14420355690850153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,51200,0.047500444783104785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,12288,0.014691554837756686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,51200,0.11455999480353461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,16384,0.019193778435389202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,16384,0.019384000036451552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,10240,0.013032889200581444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,16384,0.04394577940305074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,12288,0.015942222542232938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,12288,0.03507555524508158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,10240,0.013108444710572561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,10240,0.014118222726715935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,128,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,10240,0.031160000297758315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,8192,0.011432888607184092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,7168,0.011545778148704104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,8192,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,5120,0.019934222102165222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,8192,0.026866666144794885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,6144,0.010837333069907295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,7168,0.010718222293588849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,7168,0.011410666836632622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,5120,0.010306666294733683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,6144,0.010031999813185798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,6144,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,6144,0.02254755629433526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,4096,0.008871999879678091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,5120,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,5120,0.00904088881280687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,12288,0.015169777803950839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,5120,0.019996444384257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,4096,0.00870488915178511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,4096,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,8192,0.012787555654843649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,4096,0.0177520000272327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,3072,0.007644444704055786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,3584,0.008035555481910706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,3584,0.007744000189834171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,2560,0.005704888867007361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,3584,0.016167110866970487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,3072,0.007855999800893996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,3072,0.007366221812036302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,2048,0.004834666848182678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,3072,0.01516888870133294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,2560,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,2560,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,1536,0.004259555704063839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,7168,0.024551999237802293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,2048,0.007153777612580194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,2048,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,1024,0.003539555602603488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,2048,0.013109332985348172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,1536,0.00601244428091579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,1536,0.012147555748621622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,768,0.003165333428316646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,1024,0.006311111152172089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,3584,0.008205333517657386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,1024,0.011025778121418424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,512,0.002790222151411904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,768,0.00630044440428416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,768,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,256,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,768,0.010384889112578498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,512,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,512,0.0052915554907586836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,128,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,512,0.010063110954231685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,256,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,256,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,64,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,128,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,2560,0.014173333843549093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,2048,32,0.0021359999146726397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,128,0.007343110938866933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,65536,0.042818665504455566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,1536,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,64,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,32,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,51200,0.03407111101680332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,65536,0.05940533346600003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,2048,1024,0.0058417779703934984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,16384,0.015138667490747241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,65536,0.14336000548468694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,51200,0.047232889466815524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,51200,0.11413066916995579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,16384,0.01907999979125129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,12288,0.012486221889654795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,16384,0.018371555540296767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,256,0.00563822231358952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,16384,0.04344711038801405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,10240,0.012912888493802814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,12288,0.015496888094478183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,12288,0.03473688827620612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,128,0.005671999934646819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,8192,0.012423999607563019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,10240,0.013109332985348172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,10240,0.01346577786737018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,7168,0.011468444433477191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,8192,0.011415111521879831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,8192,0.012103111379676394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,8192,0.02660622199376424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,65536,0.05821777714623345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,7168,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,6144,0.010262221925788457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,7168,0.010694222317801582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,7168,0.023917333947287664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,51200,0.04657599992222256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,5120,0.008947555389669206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,6144,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,6144,0.021312000022994146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,4096,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,5120,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,5120,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,12288,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,5120,0.019469334019554984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,3584,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,4096,0.008610666626029545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,4096,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,3072,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,4096,0.017232000827789307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,3584,0.008268444074524773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,3584,0.007684444387753804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,10240,0.030812445614073012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,2560,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,3584,0.016158221496476066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,3072,0.00792266676823298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,3072,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,2560,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,3072,0.015077332655588785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,2560,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,1536,0.0038924444880750445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,2560,0.014135999812020196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,2048,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,6144,0.010051555931568146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,1536,0.00665155549844106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,2048,0.01314311143424776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,1024,0.003216000066863166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,1536,0.006173333360089197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,768,0.0028942223224374983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,1536,0.012087110843923358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,1024,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,1024,0.0058168888919883305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,512,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,1024,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,768,0.006073777874310811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,768,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,512,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,512,0.0053191110491752625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,512,0.010011555420027839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,2048,0.004518222063779831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,128,0.002159111201763153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,256,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,256,0.005648889061477449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,256,0.00905600024594201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,2048,0.006552888701359431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,128,0.005640000104904175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,128,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,65536,0.03300088975164626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,64,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,32,0.005187555733654234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,65536,0.057090666559007436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,768,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,65536,0.05731022357940674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,51200,0.026390223039521113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,65536,0.14316622416178384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,16384,0.012143111063374413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,51200,0.046047998799218066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,51200,0.045237332582473755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,51200,0.11381333404117161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,16384,0.018180444836616516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,16384,0.016959110895792644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,16384,0.04341422186957466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,64,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1536,128,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,10240,0.01147822207874722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,12288,0.014482667048772177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,12288,0.034279998805787824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,8192,0.010140444669458602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,10240,0.0127368892232577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,10240,0.012792888614866467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,10240,0.030807110998365615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,7168,0.009639110830095079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,8192,0.011371555427710215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,8192,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,256,0.0023333333018753263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,8192,0.02574133376280467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,6144,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,7168,0.010382222632567087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,7168,0.023355555203225877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,5120,0.008061333662933773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,6144,0.009714666340086196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,6144,0.009664889011118147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,12288,0.012950222525331708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,4096,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,6144,0.021336888273557026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,5120,0.008958222137557136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1536,32,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,3584,0.010657777388890585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,5120,0.01925244430700938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,12288,0.01407999959256914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,4096,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,4096,0.01760444376203749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,3072,0.00941066692272822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,3584,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,3584,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,3584,0.015856888559129503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,2560,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,3072,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,3072,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,3072,0.014831110835075378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,2048,0.007184888753626082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,2560,0.007520000139872233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,2560,0.006877333339717653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,2560,0.013824000126785703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,1536,0.0042239998777707415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,2048,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,2048,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,2048,0.013101333545313941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,1024,0.003536888708670934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,5120,0.00868888861603207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,1536,0.006008000009589725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,4096,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,1536,0.011997333003415002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,768,0.003167111012670729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,1024,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,1024,0.005647999958859549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,512,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,1024,0.011071999867757162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,768,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,768,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,256,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,768,0.01038933296998342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,512,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,512,0.00527555536892679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,128,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,512,0.00997155573632982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,7168,0.0103493332862854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,256,0.005098666581842634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,64,0.0021582223061058256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,256,0.009036444127559662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,1024,32,0.0021395554973019496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,128,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,65536,0.02621511121590932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,128,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,64,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,32,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,1536,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,51200,0.022099556194411382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,65536,0.0570088889863756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,65536,0.05562311410903931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,51200,0.04556355542606778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,51200,0.04457066787613762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,65536,0.14356444941626653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,12288,0.010758221977286868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,51200,0.11342133416069879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,16384,0.017661333084106445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,16384,0.016543999314308167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,10240,0.011594666375054253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,16384,0.043150222963756986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,12288,0.014110222458839417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,12288,0.014137778017255994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,8192,0.010313777459992303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,12288,0.03438933359252082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,10240,0.012839111189047495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,10240,0.012724444270133972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,7168,0.009580444130632613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,1024,128,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,8192,0.011025778121418424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,10240,0.029688000679016113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,8192,0.010743111371994019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,6144,0.008621333373917473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,8192,0.02570844524436527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,7168,0.010629333555698395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,5120,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,7168,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,6144,0.009719111025333405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,4096,0.007002666592597961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,6144,0.009330666727489894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,7168,0.023548444112141926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,5120,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,16384,0.012635555532243518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,3584,0.006721777634488211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,6144,0.021320889393488567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,5120,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,5120,0.01923466722170512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,4096,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,3584,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,4096,0.016929777132140267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,3584,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,2560,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,3584,0.015856888559129503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,3072,0.007702222300900354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,3072,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,2048,0.004577777865860197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,2560,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,3072,0.015085332923465304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,1536,0.003847111016511917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,2560,0.00673333348499404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,2560,0.013815110756291283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,2048,0.006554666492674086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,2048,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,1024,0.0032071111102898917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,1536,0.006877333339717653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,1536,0.011805333197116852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,1024,0.005647999958859549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,768,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,1024,0.01088711122671763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,3072,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,768,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,768,0.006095110956165526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,512,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,768,0.010471111370457543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,512,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,512,0.005306666509972678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,256,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,512,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,2048,0.007000000112586551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,128,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,256,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,256,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,256,0.009510222408506606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,64,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,128,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,128,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,768,32,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,1024,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,65536,0.02495644489924113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,128,0.007620444728268518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,64,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,4096,0.008348444269763099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,32,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,51200,0.021126222279336717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,65536,0.056539555390675865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,65536,0.05453600154982673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,16384,0.014460444450378418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,51200,0.04455022348297966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,65536,0.14244622654385036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,12288,0.010341333018408881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,51200,0.04388711187574598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,16384,0.01682844426896837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,51200,0.11337510744730632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,10240,0.009478222164842818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,16384,0.04260266820589701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,12288,0.014192889134089151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,12288,0.014105778601434497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,8192,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,12288,0.033547557062572904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,768,1536,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,10240,0.012558221817016602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,7168,0.007916444705592262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,10240,0.029380444023344252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,8192,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,8192,0.025643555654419795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,6144,0.006993778049945831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,7168,0.010319111247857412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,7168,0.010277333358923594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,5120,0.006771555377377405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,7168,0.022825777530670166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,6144,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,6144,0.009320889082219865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,4096,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,5120,0.008967111508051554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,5120,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,16384,0.017303999927308824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,5120,0.018971555762820773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,3584,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,4096,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,4096,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,3072,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,4096,0.016879111528396606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,3584,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,3584,0.007631111476156447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,10240,0.01276711126168569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,2560,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,3584,0.015846222639083862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,3072,0.00776800016562144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,3072,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,2048,0.004322666674852371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,8192,0.010972444381978778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,2560,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,1536,0.003871110992299186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,2560,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,2048,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,2048,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,1024,0.0031724443866146933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,2048,0.013197333448463015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,1536,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,1536,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,768,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,1536,0.011753777662913004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,6144,0.020977778567208186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,1024,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,512,0.002465777720014254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,768,0.005969777703285217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,768,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,768,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,256,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,512,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,512,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,512,0.009663110805882348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,256,0.005187555733654234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,128,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,3072,0.01479111115137736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,256,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,2560,0.013847110999955071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,128,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,64,0.0018880000958840053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,512,32,0.0018826667219400406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,128,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,65536,0.024767110745112102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,128,0.0075804442167282104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,64,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,1024,0.006311999840868845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,512,32,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,51200,0.020246222615242004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,65536,0.05383022295104133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,1024,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,65536,0.0521395570702023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,16384,0.008425777984990014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,65536,0.14174666669633654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,51200,0.039910223748948835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,51200,0.04214222232500712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,12288,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,16384,0.01648977730009291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,16384,0.01680799987581041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,10240,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,16384,0.04216444492340088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,12288,0.01383733335468504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,12288,0.013817778064144982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,8192,0.009327111144860586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,12288,0.03355200091997782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,10240,0.012549333274364471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,10240,0.012423111332787408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,7168,0.009462222456932068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,10240,0.029353777567545574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,8192,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,8192,0.025416889124446448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,6144,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,7168,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,7168,0.023342221975326538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,6144,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,6144,0.009732444253232744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,6144,0.020623111062579684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,51200,0.11241866482628717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,4096,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,5120,0.009120889008045197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,5120,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,256,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,5120,0.018914666440751817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,3584,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,4096,0.008360889222886827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,4096,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,3072,0.005631999837027655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,4096,0.017550221747822232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,3584,0.008163555628723567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,8192,0.010675555302037133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,2560,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,3584,0.015844444433848064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,3072,0.007690666450394525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,3072,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,2048,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,3072,0.015174221661355762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,2560,0.0074337778819931885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,2560,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,1536,0.003547555456558863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,2560,0.013746666411558786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,2048,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,5120,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,2048,0.006330666856633292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,1024,0.003176888864901331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,2048,0.013082666529549493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,1536,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,768,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,1536,0.011702222128709158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,1024,0.006409777949253718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,1024,0.011028444601429833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,512,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,768,0.006260444306664997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,768,0.0053004444473319584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,768,0.010045333041085137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,512,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,256,0.0021964444054497615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,3584,0.007364444434642792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,512,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,128,0.0019128889673286015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,512,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,7168,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,256,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,256,0.008983111215962304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,64,0.0018044445249769424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,128,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,256,32,0.0017991111510329777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,65536,0.014334221680959066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,128,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,64,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,1536,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,32,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,51200,0.012707555459605323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,65536,0.051743110020955406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,256,1024,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,65536,0.04896355668703715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,16384,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,65536,0.14125688870747885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,51200,0.041430221663581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,51200,0.03938311007287767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,12288,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,51200,0.11177066961924235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,16384,0.016789333687888253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,16384,0.016523554921150208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,10240,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,16384,0.04209511147605049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,12288,0.014102222190962898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,256,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,12288,0.013448889056841532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,8192,0.005649777750174205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,12288,0.03357866737577651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,10240,0.01277599980433782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,7168,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,8192,0.011095110740926532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,10240,0.029177778297000464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,8192,0.025047999289300706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,6144,0.006128888991143968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,7168,0.00998044427898195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,7168,0.023081777824295893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,6144,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,5120,0.00563733321097162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,6144,0.00980355590581894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,6144,0.02125866711139679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,5120,0.009072889056470659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,4096,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,5120,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,5120,0.01924444403913286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,10240,0.012216889195972018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,4096,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,4096,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,4096,0.016524444023768108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,8192,0.010397333237859938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,3072,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,3584,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,3584,0.0075786668393347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,2560,0.005212444398138259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,3584,0.016158221496476066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,3072,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,3072,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,2048,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,3072,0.014836443795098199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,2560,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,2560,0.006646222124497096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,1536,0.003551111039188173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,2560,0.014127999544143677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,2048,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,2048,0.006066666709052191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,1024,0.0028942223224374983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,2048,0.012774222426944308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,1536,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,1536,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,768,0.002789333255754577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,1536,0.01202133297920227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,1024,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,128,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,1024,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,3584,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,1024,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,768,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,768,0.00553688903649648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,256,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,512,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,512,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,7168,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,128,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,512,0.009720000127951304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,256,0.005140444470776452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,256,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,64,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,256,0.008956444760163626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,32,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,128,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,128,0.004950222041871813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,65536,0.010504888991514841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,128,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,16384,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,64,0.0047564444442590075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,12288,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,128,32,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,10240,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,65536,0.05065066615740458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,51200,0.04093599981731839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,8192,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,16384,0.016487111647923786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,7168,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,6144,0.005200888961553574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,12288,0.01369599997997284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,5120,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,128,512,0.002464888824356927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,10240,0.012048888537618848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,8192,0.010684444672531553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,4096,0.004610666798220741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,3584,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,3072,0.004862222406599256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,6144,0.009034666750166152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,768,0.010403555300500657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,5120,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,2560,0.004975999808973736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,3584,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,2048,0.004613333278232151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,1536,0.003519111209445529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,3072,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,1024,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,2560,0.006486222147941589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,768,0.002593777866827117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,2048,0.006059555543793573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,51200,0.011884444289737277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,1536,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,256,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,1024,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,128,0.0018622221218215094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,768,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,64,0.001820444439848264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,512,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,256,0.00508266646001074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,128,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,65536,0.009728888670603434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,64,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,51200,0.008691555923885768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,32,0.005071111023426056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,7168,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,16384,0.006805333412355847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,12288,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,64,4096,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,65536,0.050632887416415744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,10240,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,8192,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,16384,0.016532444291644625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,7168,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,12288,0.013780444032616086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,6144,0.005057777795526716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,10240,0.012095111111799875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,5120,0.004583111239804162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,8192,0.010509333676762052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,4096,0.004527111020353106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,7168,0.009719111025333405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,3584,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,6144,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,3072,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,5120,0.008290666672918532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,32,0.0018151110659042995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,4096,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,2048,0.004537777768241035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,3584,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,1536,0.0035502223504914176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,3072,0.00701422244310379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,1024,0.0029004443850782183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,2560,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,768,0.0028071111689011254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,2048,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,512,0.002505777817633417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,256,0.002116444417172008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,1536,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,128,0.001869333287080129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,1024,0.005640000104904175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,51200,0.04076888826158311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,64,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,768,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,32,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,512,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,256,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,128,0.00461777796347936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,64,512,0.002463999928699599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,64,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,160,32,32,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,160,32,2560,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,16384,0.3391039901309543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,12288,0.25914398829142254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,16384,0.26832977930704754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,16384,0.13456888993581137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,10240,0.21368088987138537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,12288,0.19974933730231392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,51200,1.0421039793226454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,8192,0.17892800437079537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,12288,0.10154844654930963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,10240,0.16718133290608725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,10240,0.08481333653132121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,7168,0.15449510680304632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,8192,0.13824978139665392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,6144,0.13268978065914577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,7168,0.12483110692765977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,51200,0.4083840052286784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,8192,0.06842488712734647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,7168,0.061780446105533175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,6144,0.05352088809013367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,5120,0.11244977845085992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,5120,0.08828444613350762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,4096,0.09362755881415473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,3584,0.0817999972237481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,5120,0.047206223011016846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,4096,0.03917510973082648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,3072,0.07131733496983846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,4096,0.0731395549244351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,51200,0.8296648661295573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,3584,0.03577955563863119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,2560,0.059826665454440646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,3584,0.06326933039559259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,3072,0.030896888838873968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,3072,0.05560533205668131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,2048,0.049264887968699135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,2560,0.02723466687732273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,2560,0.047925333182017006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,1536,0.03815644317203098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,2048,0.02404444416364034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,1024,0.029024001624849107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,1536,0.018651554981867473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,1536,0.0319217774603102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,1024,0.024899555577172175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,1024,0.015503999259736804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,512,0.014767110347747803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,768,0.014072888427310519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,256,0.00907199995385276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,768,0.03052266769938999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,512,0.013242666920026144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,128,0.007891555627187094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,512,0.022057778305477563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,6144,0.10203111171722412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,256,0.015844444433848064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,64,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,32,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,128,0.011415111521879831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,128,0.01477599971824222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,32,0.01553066737122006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,2048,0.04035910964012146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,65536,768,0.021380444367726643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,65536,256,0.011639110743999481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,65536,0.8364035818311902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,65536,1.1335698233710396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,51200,0.8878044552273221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,51200,0.668143113454183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,65536,64,0.01456622282663981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,12288,0.21742488278283012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,65536,0.4040035671657986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,12288,0.15965955787234837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,16384,0.211937771903144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,12288,0.08038666513231066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,10240,0.19730577203962538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,8192,0.16095111105177137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,16384,0.1077386670642429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,51200,0.3123857710096571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,10240,0.1361022260453966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,7168,0.14082488748762342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,8192,0.11246221595340306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,6144,0.12165689468383789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,8192,0.05591999822192722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,7168,0.09835377666685317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,7168,0.04955110947291056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,5120,0.1023484468460083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,16384,0.2865235540601942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,6144,0.04342222213745117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,6144,0.08285599946975708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,4096,0.08263822396596272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,5120,0.0373564428753323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,3584,0.07376533084445529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,3072,0.06457600328657362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,4096,0.057357331116994224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,3584,0.029579556650585596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,4096,0.03256533212131924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,3584,0.05079466766781277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,2560,0.05406666795412699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,2048,0.04373066624005636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,3072,0.0258577780591117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,3072,0.044962667756610446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,1536,0.0340577761332194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,2560,0.023247111174795363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,2048,0.0199297782447603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,1024,0.024392000503010217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,2048,0.033005333609051175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,1536,0.01628977722591824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,10240,0.06763377454545763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,1536,0.02687377731005351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,512,0.013269333375824822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,1024,0.01386755539311303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,1024,0.0207022229830424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,256,0.008641777767075432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,512,0.011368888947698804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,512,0.018411555223994784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,128,0.006125333408514659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,256,0.01036622209681405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,5120,0.0713262226846483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,256,0.013384000294738345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,64,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,128,0.010382222632567087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,32,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,128,0.012425777812798819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,64,0.012498666842778524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,32,0.013287111288971372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,65536,0.3395564556121826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,2560,0.039056890540652804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,51200,768,0.01850311126973894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,65536,0.3001520103878445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,16384,0.09017599953545465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,51200,0.2655751175350613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,51200,768,0.012787555654843649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,12288,0.06942044364081489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,51200,768,0.025398222936524287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,65536,0.13771377669440374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,51200,0.1086462206310696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,16384,0.03968088825543722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,51200,0.23254399829440645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,10240,0.05698577562967936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,16384,0.07776977618535359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,12288,0.031522668070263334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,10240,0.02743911080890232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,7168,0.04154488775465224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,8192,0.02310933338271247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,8192,0.043199112017949425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,12288,0.059064891603257924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,6144,0.03572266631656223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,7168,0.02215733296341366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,5120,0.030551112360424463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,7168,0.037512888511021934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,4096,0.026183111800087824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,6144,0.03342133429315355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,5120,0.016600888636377122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,3584,0.022652443912294176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,5120,0.02871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,4096,0.014815110299322339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,4096,0.024179556303554114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,3072,0.01994400057527754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,3584,0.013753777576817406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,2560,0.016508445143699646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,3584,0.021733333667119343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,3072,0.01272266689274046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,2048,0.013746666411558786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,3072,0.01936533384852939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,2560,0.012006222373909421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,1536,0.011018666956159802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,2560,0.017076444294717576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,2048,0.010888000329335531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,1024,0.008613333106040955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,1536,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,8192,0.04814399944411384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,10240,0.05063555637995402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,768,0.007038222418891058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,1536,0.012783110969596438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,512,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,1024,0.008003555238246918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,1024,0.010478222535716163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,256,0.0035439998739295532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,768,0.007674666742483775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,6144,0.01886311173439026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,128,0.003170666595300039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,768,0.012101333174440594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,64,0.0029120000286234748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,512,0.009229333036475712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,16384,32,0.0031964443624019623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,128,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,256,0.007476444045702617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,128,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,64,0.007252444823582967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,65536,0.27384887801276314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,32,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,51200,0.21411911646525064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,65536,0.21326400174034965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,16384,0.07202221949895223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,65536,0.1257715490129259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,16384,2048,0.014831999937693277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,12288,0.05652088589138455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,51200,0.09976266490088569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,16384,0.056687997447119824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,16384,0.037314666642083064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,10240,0.0464933349026574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,8192,0.03939022289382087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,12288,0.04385333259900411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,512,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,10240,0.025456888808144465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,7168,0.03398844599723816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,10240,0.03746755586730109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,16384,256,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,8192,0.02166044380929735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,6144,0.02921600143114726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,8192,0.03199822372860379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,7168,0.019720888800091214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,7168,0.02834133307139079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,6144,0.017893332574102614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,4096,0.02158222264713711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,5120,0.016487111647923786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,3584,0.018586667047606576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,5120,0.021300445000330608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,4096,0.01442399952146742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,3072,0.016341333587964375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,4096,0.018486221631368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,3584,0.013474666410022311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,3584,0.016584888100624084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,2560,0.013779555757840475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,3072,0.012404444317022959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,2048,0.011763555308183035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,51200,0.16481867101457384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,2560,0.011328000161382886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,1536,0.009744000103738572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,2560,0.013737777868906656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,2048,0.009752000371615091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,1024,0.007703999678293864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,2048,0.012517333030700684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,1536,0.008359111017651027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,1536,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,768,0.006670222100284364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,5120,0.025172443853484258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,1024,0.007672000262472365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,1024,0.008303110798199972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,512,0.005043555465009478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,6144,0.025232000483406916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,768,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,256,0.003866666720973121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,512,0.007159111400445302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,768,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,128,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,512,0.008012444608741337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,64,0.0028168888141711554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,256,0.006647999915811751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,12288,32,0.002834666727317704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,128,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,256,0.006870222174459034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,128,0.006670222100284364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,64,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,32,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,65536,0.2659866544935438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,12288,12288,0.029843555556403265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,65536,0.2014088895585802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,51200,0.2086026668548584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,16384,0.06946755780114068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,65536,0.12317511770460342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,12288,0.054153776831097074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,51200,0.15776356061299643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,51200,0.09800355301962958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,10240,0.04517599940299988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,12288,0.028660443094041612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,12288,0.04192089041074117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,8192,0.03792000148031447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,10240,0.024490666058328416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,7168,0.03306577934159173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,8192,0.021145777569876775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,8192,0.030930668115615845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,6144,0.028961777687072754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,7168,0.019679110911157396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,7168,0.026800000005298193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,12288,3072,0.015111999379263984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,5120,0.024819556209776137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,6144,0.017514665921529133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,6144,0.0236124445994695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,4096,0.020815999971495736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,5120,0.02055911057525211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,3584,0.01811377704143524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,4096,0.014144000079896716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,4096,0.01763911048571269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,3072,0.016192888220151264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,16384,0.035875555541780256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,16384,0.05416088965204027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,2560,0.013724444641007317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,3584,0.013394667042626275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,2048,0.011700444751315646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,3072,0.012067555553383298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,3584,0.01614755557643043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,1536,0.009648000200589498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,2560,0.010458666417333813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,3072,0.014864888456132678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,1024,0.007644444704055786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,2560,0.013227555486891003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,10240,0.03609244359864129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,2048,0.0086986662613021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,768,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,1536,0.008290666672918532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,2048,0.011757333245542316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,512,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,1536,0.010108444425794814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,768,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,256,0.003847111016511917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,1024,0.008162666526105668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,768,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,512,0.006907555378145642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,128,0.0029208889851967492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,64,0.002830222249031067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,256,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,10240,32,0.002833777831660377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,256,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,128,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,5120,0.01553422212600708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,128,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,64,0.006364444477690591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,65536,0.1789217789967855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,32,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,51200,0.14412889215681288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,16384,0.0510062211089664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,65536,0.07440000110202365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,12288,0.04039288891686334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,51200,0.12370577123430039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,51200,0.060106668207380504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,16384,0.023743111226293776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,16384,0.04355644517474704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,10240,0.03321511215633816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,8192,0.02769244379467434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,12288,0.019056888090239633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,12288,0.034049779176712036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,10240,1024,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,10240,0.02920266654756334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,7168,0.02446488870514764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,8192,0.014930667148696052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,8192,0.025010666913456384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,6144,0.021387555532985266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,10240,512,0.007592000067234039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,7168,0.013850666582584381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,5120,0.018340445227093168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,7168,0.021664889322386846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,6144,0.012735111018021902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,4096,0.016009777784347534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,6144,0.01940711173746321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,3584,0.013400889105266996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,5120,0.011795555551846823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,5120,0.016992888516849942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,4096,0.010390222072601318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,3072,0.012325333224402534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,4096,0.015030221806632148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,2560,0.010476444330480365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,3584,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,65536,0.1588382191128201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,3072,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,2048,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,3072,0.012815111213260226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,2560,0.008127999802430471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,1536,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,2048,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,2560,0.011751111182901593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,1024,0.005713777823580637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,2048,0.01035199976629681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,1536,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,768,0.004536889079544279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,10240,0.01687733332316081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,1536,0.008812444077597724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,512,0.0039022221333450745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,768,0.006502222269773483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,1024,0.007324444750944774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,256,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,512,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,768,0.008284444610277811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,128,0.0025519999778932994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,256,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,64,0.0024666666156715816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,256,0.006145777801672618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,8192,32,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,128,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,64,0.006230222268237009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,65536,0.16386044025421143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,32,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,3584,0.013766222529941134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,51200,0.1307777828640408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,65536,0.15349600050184461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,65536,0.07148089011510213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,16384,0.048992888795004956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,51200,0.1214631133609348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,1024,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,12288,0.03871555460823907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,51200,0.05779555771085951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,10240,0.03233155608177185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,16384,0.022508444057570562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,16384,0.04241333405176798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,8192,0.027479999595218237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,12288,0.01832799944612715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,12288,0.0329733325375451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,10240,0.01647555496957567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,7168,0.023895111348893907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,6144,0.02087733315096961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,10240,0.028163555595609877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,8192,0.014592889282438489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,5120,0.017644445101420086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,7168,0.013408000270525614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,8192,0.023688889212078516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,4096,0.015256888336605497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,7168,0.020979555116759405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,5120,0.011360000405046674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,6144,0.019154666198624503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,3584,0.01309333327743742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,5120,0.01683022247420417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,4096,0.010023111270533668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,3072,0.011894221935007306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,4096,0.014779556128713818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,3584,0.008981333010726506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,2560,0.010038221875826517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,3072,0.008271111382378472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,2048,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,2560,0.007872888611422645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,3072,0.012498666842778524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,1536,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,2560,0.011472889118724398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,2048,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,1024,0.00566133318675889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,2048,0.009749333063761393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,1536,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,768,0.004528000122971005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,1536,0.008347555167145198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,8192,512,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,1024,0.00656088896923595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,512,0.003671110918124517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,8192,128,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,1024,0.007255111303594377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,768,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,256,0.0031511110977994073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,768,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,512,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,128,0.0025333333760499954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,6144,0.012420444852775998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,512,0.006689777804745569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,64,0.0025111111915773815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,7168,32,0.0028560000161329904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,128,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,256,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,128,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,64,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,65536,0.1482506725523207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,32,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,7168,3584,0.013720889058378009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,51200,0.12210133340623643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,65536,0.11589866214328343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,65536,0.06908888949288262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,12288,0.03360266817940606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,51200,0.05560355716281467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,16384,0.02167911165290409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,51200,0.09161421987745498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,10240,0.029158221350775823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,16384,0.03285422258906894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,12288,0.01773866679933336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,8192,0.02304977840847439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,12288,0.025362667110231187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,10240,0.01588888963063558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,7168,0.02010311186313629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,10240,0.021874666213989258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,8192,0.013919111755159167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,6144,0.01754666699303521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,8192,0.018930666976504855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,7168,0.013186666700575085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,7168,256,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,6144,0.012062222593360476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,7168,0.016957332690556843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,4096,0.012830221818553077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,6144,0.015544889701737298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,3584,0.011301333705584208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,5120,0.01069777790043089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,4096,0.009263111485375298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,5120,0.013776000175211163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,3584,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,2560,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,16384,0.04529066549407112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,3584,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,2048,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,3072,0.01071911139620675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,2560,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,1536,0.006573333508438534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,2048,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,2048,0.008346666892369589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,1024,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,1536,0.006923555499977536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,1536,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,768,0.0038844446341196695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,1024,0.0064177778032090925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,1024,0.006573333508438534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,5120,0.015207111835479736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,512,0.0032195556494924757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,768,0.006254222244024277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,256,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,768,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,512,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,3072,0.010218666659461128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,128,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,256,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,256,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,64,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,6144,32,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,3072,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,128,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,128,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,65536,0.11806844340430366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,64,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,2560,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,32,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,51200,0.09446489148669773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,16384,0.037132445308897234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,65536,0.11434045102861191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,65536,0.06642311149173312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,12288,0.03254666593339708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,51200,0.053730666637420654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,16384,0.020894222789340552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,10240,0.027616888284683228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,16384,0.03161244591077169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,12288,0.017432888348897297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,12288,0.024596444434589807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,8192,0.023044443792766992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,7168,0.02010933392577701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,10240,0.02146400014559428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,6144,512,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,8192,0.013606222139464485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,6144,0.017477333545684814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,7168,0.01275644451379776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,5120,0.014867555763986377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,7168,0.016600888636377122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,6144,0.011651555697123209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,6144,0.01519466605451372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,4096,0.0125937776433097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,5120,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,5120,0.01366222235891554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,3584,0.01110311100880305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,4096,0.008672888908121321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,4096,0.012401777837011548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,3072,0.010015111002657149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,3584,0.00832177781396442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,51200,0.08849422136942546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,2560,0.008621333373917473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,3072,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,3072,0.010423111418883005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,2560,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,2560,0.008976889153321585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,8192,0.018626666731304593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,1536,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,2048,0.007223111059930589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,6144,4096,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,2048,0.008055999875068665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,1024,0.004858666823969947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,1536,0.006761777732107375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,768,0.0038444445365005066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,1024,0.0063982220987478895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,1024,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,512,0.003208000005947219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,768,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,768,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,256,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,512,0.006191111273235745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,3584,0.011156444748242697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,128,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,256,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,256,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,64,0.0021582223061058256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,128,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,2048,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,5120,32,0.0023093333260880574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,128,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,65536,0.09359466367297703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,64,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,32,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,51200,0.07639733288023207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,65536,0.06429155667622884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,65536,0.10605244504080878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,1536,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,16384,0.03314044409328037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,51200,0.08389066987567478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,51200,0.05147200160556369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,16384,0.02028888960679372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,10240,0.02091911103990343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,16384,0.029657777812745836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,12288,0.016765332884258695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,5120,512,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,10240,0.015115555789735583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,10240,0.02070755594306522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,7168,0.015398222539159985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,8192,0.013099555340078143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,8192,0.017793777916166518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,6144,0.013438222308953604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,5120,10240,0.015499555402331881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,7168,0.012105777859687805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,5120,0.011752888560295105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,7168,0.016303999556435477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,6144,0.010942222343550788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,4096,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,5120,0.009711110757456886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,3584,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,5120,0.013244444297419654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,4096,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,3072,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,4096,0.011515555282433828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,3584,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,3584,0.010738666686746808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,2560,0.006792000184456508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,3072,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,12288,0.02587466604179806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,2048,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,3072,0.009754666851626502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,1536,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,2560,0.007350222104125553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,2560,0.008634666601816813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,8192,0.017477333545684814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,1024,0.0038702223036024305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,2048,0.00794222205877304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,12288,0.023585778143670823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,768,0.003532444437344869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,1536,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,1024,0.00656444455186526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,768,0.006224000205596288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,1024,0.006649777707126405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,256,0.002536888958679305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,768,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,512,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,512,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,128,0.0024871110088295406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,256,0.005646222167544895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,6144,0.014782221780882942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,64,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,256,0.005623111294375525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,128,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,32,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,128,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,65536,0.08576444122526382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,64,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,4096,32,0.00554044461912579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,51200,0.07017511129379272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,65536,0.06311111317740546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,65536,0.10411377747853597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,16384,0.03176533182462057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,51200,0.050345778465270996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,1536,0.006754666566848755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,51200,0.08151555723614164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,16384,0.019952888290087383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,4096,512,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,10240,0.021683555510309007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,16384,0.02944533361328973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,12288,0.016565332810084026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,8192,0.01657777859105004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,12288,0.023415999280081853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,10240,0.01479111115137736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,7168,0.014475555883513557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,10240,0.020317332612143624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,8192,0.01276088919904497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,6144,0.012852444416946836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,8192,0.017536888519922893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,7168,0.011698666546079846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,5120,0.011115555961926779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,7168,0.015882665912310284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,6144,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,4096,0.009877333210574256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,6144,0.014493332968817817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,5120,0.009381333159075843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,3584,0.008667555948098501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,5120,0.013069333301650153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,4096,0.008602666358153025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,3072,0.007920888562997183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,4096,0.011340444286664328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,2560,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,3584,0.010038221875826517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,4096,2048,0.00720177756415473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,2048,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,3072,0.007732444339328342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,3072,0.009152000149091085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,2560,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,12288,0.02509333358870612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,2560,0.00867911097076204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,2048,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,1024,0.0038364442686239877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,768,0.003338666632771492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,2048,0.00793422261873881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,1536,0.0068400001360310455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,512,0.0028924445311228433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,1024,0.006360888895061281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,1536,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,256,0.002518222149875429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,768,0.006115555349323485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,1024,0.00646577775478363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,128,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,512,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,256,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,64,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,256,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,32,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,128,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,65536,0.07452799876530965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,64,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,3584,0.008275555239783393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,32,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,51200,0.06058400207095676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,65536,0.062168002128601074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,65536,0.06580711073345609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,16384,0.024439111351966858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3584,1536,0.00480000012450748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,12288,0.020488889680968392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,51200,0.04926400052176582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,16384,0.01965777741538154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,10240,0.01716088917520311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,16384,0.019744000501102872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,12288,0.016307555966907077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,8192,0.013965333501497904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,768,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,10240,0.014525334040323893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,7168,0.012527999778588613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,10240,0.0144177774588267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3584,128,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,8192,0.012312889099121094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,6144,0.011189333266682096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,7168,0.011343999869293638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,5120,0.009827555881606208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,6144,0.010284444524182213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,6144,0.011008888483047485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,4096,0.008615111311276754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,5120,0.009343110852771336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,5120,0.009905777871608734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,3584,0.007825777762466008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,4096,0.008396444221337637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,4096,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,51200,0.05115466647677951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,3584,0.008007999923494127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,3584,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,3072,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,12288,0.016144000821643405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,2560,0.006512889017661412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,3072,0.007373332977294922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,2048,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3584,512,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,2560,0.007558222446176741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,8192,0.012681778106424542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,1536,0.004546666724814309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,2048,0.0071928890215026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,2560,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,1024,0.003563555578390757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,2048,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,7168,0.01184177812602785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,768,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,1536,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,512,0.0028320000403457214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,1024,0.006317333214812809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,1536,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,256,0.0025315555847353404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,1024,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,512,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,128,0.002159111201763153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,512,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,256,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,64,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,256,0.004979555391603046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,3072,0.007190222541491191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,3072,32,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,128,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,65536,0.06518666611777411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,128,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,64,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,51200,0.05262133479118347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,32,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,16384,0.02312799957063463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,65536,0.060192889637417264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,65536,0.06288799974653456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,51200,0.0481315553188324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,12288,0.018920888503392536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,51200,0.04923289020856222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,16384,0.019472888774342008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,10240,0.01646844380431705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,16384,0.019281778070661757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,12288,0.016173332929611206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,8192,0.013400000002649097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,12288,0.015671110815472074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,10240,0.014152889450391134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,7168,0.012078222301271228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,10240,0.013957333233621387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,8192,0.011709333293967776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,3072,768,0.006215999937719769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,6144,0.011014222270912595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,5120,0.009630222287442949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,3072,768,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,7168,0.010768888725174798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,4096,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,6144,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,7168,0.011385777758227454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,3584,0.0075804442167282104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,6144,0.01015911085738076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,5120,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,3072,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,5120,0.009556444154845344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,2560,0.006188444379303191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,3584,0.008023111356629265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,4096,0.008120888637171851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,3072,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,2048,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,3584,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,3072,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,2560,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,2560,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,2048,0.007335111498832703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,1024,0.0034924443397257063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,1536,0.006847111301289664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,2048,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,1536,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,1024,0.006454222318198945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,512,0.0028071111689011254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,1024,0.005646222167544895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,768,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,768,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,512,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,512,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,128,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,8192,0.012379555238617791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,256,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,64,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,256,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,32,0.0021280000607172647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,128,0.005640888793600931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2560,4096,0.008394666843944127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,128,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,65536,0.05391377872890896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,64,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,51200,0.043327109681235425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2560,32,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,1536,0.004211555752489301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,65536,0.059044447210099965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,16384,0.018263111511866253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,65536,0.06030133035447863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,51200,0.04684355523851183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,51200,0.04764266808827718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,16384,0.018948444061809115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,10240,0.013060444758998023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,12288,0.015170666906568738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,12288,0.015498666299713982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,256,0.0025351111673646504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,10240,0.01314399970902337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,10240,0.013487999637921652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,7168,0.011361777782440186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,8192,0.011415111521879831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,8192,0.011787555283970304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,6144,0.010696888797812991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,7168,0.01074488874938753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,7168,0.011066666907734342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,5120,0.009492444495360056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,6144,0.009862222605281407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,6144,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,4096,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,5120,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,5120,0.008714666797055138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,12288,0.014768888552983602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,3584,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2560,768,0.003141333245568805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,4096,0.008387555678685507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,4096,0.008013333711359236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,16384,0.0185244447655148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,3584,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,2560,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,3584,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,3072,0.00775911079512702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,2048,0.005111111121045219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,8192,0.012501333322789935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,2560,0.00758133331934611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,3072,0.007350222104125553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,2048,0.00720088928937912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,1024,0.003559111307064692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,1536,0.006717333363162146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,2048,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,768,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,1024,0.006314666734801398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,1024,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,512,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,768,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,512,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,256,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,3072,0.006670222100284364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,256,0.0058151111006736755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,128,0.0021520000365045336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,256,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,64,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,2048,128,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,32,0.0021280000607172647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,2048,1536,0.004383111165629493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,128,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,2560,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,65536,0.04369422131114536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,64,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,51200,0.034655110703574285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,1536,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,32,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,65536,0.05746489100986057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,16384,0.015509333875444202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,65536,0.05786310964160495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,12288,0.013172444370057849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,51200,0.046374221642812095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,51200,0.0456008878019121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,10240,0.01200711147652732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,12288,0.01408088869518704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,16384,0.018406222263971966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,8192,0.010899555351999072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,12288,0.015140444040298462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,10240,0.012788444757461548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,7168,0.011955555942323474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,512,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,10240,0.013372444444232516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,6144,0.011410666836632622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,8192,0.011363555159833698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,7168,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,5120,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,6144,0.009703999592198266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,7168,0.01034755590889189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,4096,0.007761778102980719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,6144,0.009663110805882348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,5120,0.00906044410334693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,3584,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,4096,0.00834222220712238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,3072,0.00618488879667388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,3584,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,3584,0.007689778175618913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,2560,0.005186666631036335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,3072,0.007669332954618666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,3072,0.007253333098358578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,2048,0.004534222185611725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,2560,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,16384,0.01790577835506863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,2048,768,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,1536,0.003908444609906939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,2048,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,2048,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,1024,0.0033244445092148254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,1536,0.006657777975002925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,8192,0.011351111034552256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,1536,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,1024,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,1024,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,512,0.0025439999169773525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,768,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,768,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,5120,0.0087333329849773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,512,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,4096,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,512,0.0052977779673205475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,128,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,256,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,64,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1536,128,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,32,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,128,0.005076444397370021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,65536,0.031911111540264554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,2560,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,64,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,51200,0.025807999902301367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,32,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,65536,0.057051552666558154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,65536,0.05486844314469231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,768,0.002926222152180142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,16384,0.012950222525331708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,12288,0.011365333365069495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,51200,0.04353511002328661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,16384,0.016719111137919955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,10240,0.010402667025725046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,16384,0.017834666702482436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1536,256,0.0023519999037186303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,12288,0.0147386673423979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,8192,0.009488888912730748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,10240,0.012624888784355588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,7168,0.008983111215962304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,10240,0.012433778080675336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,8192,0.011130666567219628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1536,256,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,8192,0.011006222003036074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,7168,0.01036888857682546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,5120,0.008577778107590145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,7168,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,6144,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,4096,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,5120,0.009175111022260454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,3584,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,5120,0.008616888688670265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,4096,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,3072,0.005671999934646819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,4096,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,3584,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,51200,0.04605866803063286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,2560,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,3584,0.0074142225914531285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,2048,0.004555555681387584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,3072,0.007104888558387756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,2560,0.007240888973077138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,1536,0.003842666745185852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,12288,0.013928888572586907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,2560,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,1024,0.0031679999083280563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,2048,0.006411555740568373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,1536,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,768,0.0028719999310043124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,1536,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,1024,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,6144,0.00831733312871721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,512,0.002530666689078013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,1024,0.00573244442542394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,768,0.005993777679072485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,256,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,768,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,128,0.002120888895458645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,6144,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,256,0.005659555395444234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,64,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,128,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,1024,32,0.0018995555324686898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,65536,0.02738222148683336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,128,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,64,0.004947555561860402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,3072,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,32,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,51200,0.021838222940762837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,65536,0.05642133288913303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,16384,0.010858666565683154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,65536,0.053957333167394005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,12288,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,51200,0.04572622312439812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,1024,2048,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,10240,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,51200,0.042958223157458834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,16384,0.016858667135238647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,12288,0.013935999737845527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,8192,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,12288,0.013786666923099093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,10240,0.012509333590666452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,10240,0.012168000141779581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,7168,0.00961688905954361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,8192,0.011048888994587792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,6144,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,8192,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,7168,0.01036888857682546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,512,0.0052986666560173035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,5120,0.0081395556529363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,7168,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,6144,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,1024,256,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,4096,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,3584,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,5120,0.009003555609120263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,4096,0.008271111382378472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,5120,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,3072,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,3584,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,2560,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,3584,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,3072,0.0075795559419526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,2048,0.004491555607981152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,3072,0.0070364442136552595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,2560,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,16384,0.016735111673672993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,1536,0.003842666745185852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,2560,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,2048,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,1024,0.0032133333798911837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,2048,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,1536,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,768,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,1024,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,1536,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,512,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,1024,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,768,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,512,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,128,0.0021173333128293357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,512,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,256,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,64,0.0018613334331247543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,6144,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,256,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,768,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,128,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,128,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,4096,0.00794222205877304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,51200,0.017857778403494094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,64,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,768,32,0.004946666873163647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,65536,0.05603910817040337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,16384,0.011738667057620155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,65536,0.05276088913281759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,51200,0.04257066713439094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,12288,0.009755555954244401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,16384,0.016514667206340365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,51200,0.04244622257020739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,10240,0.009099555512269339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,16384,0.016534222496880423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,8192,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,10240,0.012552888856993781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,10240,0.012063110868136087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,7168,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,8192,0.011065777805116443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,6144,0.006722666737106111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,8192,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,7168,0.010399111443095738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,5120,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,7168,0.010209778116808997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,6144,0.009664889011118147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,4096,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,6144,0.009367999931176504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,65536,0.022151110900772944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,5120,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,3584,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,5120,0.008622222476535374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,4096,0.00831200016869439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,3072,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,3584,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,2560,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,3072,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,3072,0.007056000332037608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,2048,0.004237333519591225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,2560,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,2560,0.006948444578382704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,12288,0.014108444253603617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,1536,0.003850666599141227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,12288,0.01370133293999566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,2048,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,1024,0.003152888889114062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,768,768,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,2048,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,768,0.002784888984428512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,1536,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,1024,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,1536,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,512,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,1024,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,256,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,768,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,512,0.005843555761708154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,128,0.0019253332995706135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,512,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,256,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,64,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,256,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,512,32,0.0017955555684036678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,128,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,128,0.004944889081848992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,4096,0.008058666355080074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,3584,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,64,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,51200,0.01588711142539978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,512,32,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,16384,0.009491555392742157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,65536,0.05179199907514784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,51200,0.039699554443359375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,12288,0.011359111302428775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,16384,0.016535111599498324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,16384,0.016433777080641854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,10240,0.010120888551076254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,12288,0.013800889253616333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,512,768,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,8192,0.009008888569143083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,10240,0.012431999875439538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,10240,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,7168,0.008347555167145198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,8192,0.011022222538789114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,8192,0.010423999693658618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,6144,0.007756444315115611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,7168,0.01035822182893753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,65536,0.018403554956118267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,5120,0.0069955554273393415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,7168,0.009703999592198266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,6144,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,4096,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,65536,0.04892355534765455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,6144,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,3584,0.005979555762476391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,5120,0.008632000121805403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,4096,0.008329778081840938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,3072,0.005370666583379109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,51200,0.04171466827392578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,4096,0.007899555895063613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,3584,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,2560,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,3072,0.007663110891977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,3584,0.0075937774446275495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,2048,0.0042257776690853965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,2560,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,3072,0.0070879997478591064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,1536,0.003565333369705412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,2560,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,2048,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,1024,0.00314666661951277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,2048,0.00645955569214291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,1536,0.006593777901596493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,1024,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,512,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,768,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,1024,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,256,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,768,0.005422222117582957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,512,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,128,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,512,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,256,0.00566133318675889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,5120,0.008970666262838576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,256,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,256,128,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,32,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,128,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,65536,0.014582221706708273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,64,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,51200,0.012701333396964602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,65536,0.048919111490249634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,65536,0.049197332726584546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,12288,0.013743999931547375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,16384,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,51200,0.03942311141226027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,12288,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,51200,0.039472887913386025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,768,0.002798222212327851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,16384,0.016358221570650738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,10240,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,1536,0.006051555689838197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,8192,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,12288,0.013802666631009845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,10240,0.012436444560686747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,10240,0.011774222056070963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,8192,0.011063111325105032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,8192,0.010038221875826517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,6144,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,256,64,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,7168,0.009723555710580613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,6144,0.009639999932712978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,5120,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,5120,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,16384,0.01647111111217075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,4096,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,5120,0.008861333131790161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,12288,0.013473778135246701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,3584,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,7168,0.006311111152172089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,4096,0.008294222255547842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,3072,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,4096,0.007889777421951294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,3584,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,7168,0.01038577738735411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,2560,0.004855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,3584,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,3072,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,3072,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,2560,0.006350222147173352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,1536,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,256,32,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,2048,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,6144,0.009120889008045197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,1024,0.002895111011134254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,2048,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,768,0.0028044444819291434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,1536,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,1024,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,512,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,1536,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,768,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,256,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,768,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,512,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,128,0.0018648888087934917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,512,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,256,0.005632888939645555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,64,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,256,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,32,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,128,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,65536,0.018231999542978074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,128,0.004581333448489507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,51200,0.015835555063353646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,16384,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,64,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,12288,0.006223111102978389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,128,2048,0.0042444442709287005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,32,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,10240,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,129,128,2560,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,8192,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,51200,0.03935111231274075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,7168,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,6144,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,16384,0.015867556134859722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,5120,0.006325333482689328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,12288,0.013413333230548434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,10240,0.011745777395036487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,4096,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,8192,0.010135999984211391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,3584,0.005423111220200856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,7168,0.009502222140630087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,3072,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,2560,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,6144,0.00926577796538671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,2048,0.004595555365085602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,128,1024,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,4096,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,1536,0.0035253332720862497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,3584,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,1024,0.003031111011902491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,768,0.002879111096262932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,2560,0.006721777634488211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,512,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,2048,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,256,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,1536,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,1024,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,768,0.005444444302055571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,64,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,512,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,32,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,256,0.005151111218664381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,128,0.004584889031118817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,65536,0.04890222350756327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,51200,0.017902221944597032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,64,0.004954666727119022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,16384,0.009054222040706212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,32,0.00489688871635331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,12288,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,65536,0.048955556419160634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,10240,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,51200,0.03941866755485535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,8192,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,5120,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,7168,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,64,3072,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,10240,0.011706666813956367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,6144,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,8192,0.010299555957317352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,5120,0.006301333506902059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,7168,0.009723555710580613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,64,128,0.0018275556051068837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,4096,0.005648889061477449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,3584,0.005447111195988125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,6144,0.008905777500735389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,3072,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,5120,0.008193777667151557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,2560,0.004943111290534337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,4096,0.007691555553012424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,2048,0.004601777841647466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,3072,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,1536,0.0035662220584021676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,2560,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,1024,0.0028808888875775864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,2048,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,768,0.002554666664865282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,16384,0.016188444362746347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,1536,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,512,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,1024,0.00543200017677413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,256,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,12288,0.013463999662134381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,128,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,768,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,32,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,512,0.005105777747101254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,256,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,128,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,64,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,32,0.0046302220887608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,129,32,65536,0.018986667195955913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,129,32,3584,0.007153777612580194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,16384,0.37315911716885036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,16384,0.205004440413581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,51200,0.6229928864373101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,51200,1.1552907096015084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,12288,0.2812817891438802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,16384,0.13496000236935085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,12288,0.15238489045037165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,10240,0.23701956537034777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,51200,0.39097245534261066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,10240,0.13474755816989475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,12288,0.10161688592698838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,8192,0.2026533285776774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,10240,0.08535289102130467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,8192,0.1083573301633199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,7168,0.17092266347673205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,8192,0.07071555323070951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,16384,0.2972017659081353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,7168,0.09160088830524021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,6144,0.14767200416988796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,10240,0.18569689326816133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,7168,0.06138844622506035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,6144,0.07993600103590223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,8192,0.1548373301823934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,12288,0.22518489095899794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,5120,0.12313600381215413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,6144,0.054140445258882314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,7168,0.1333448886871338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,5120,0.06729422012964885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,5120,0.04649600055482653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,4096,0.05545333358976576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,3584,0.088992887073093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,6144,0.11660267247094049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,4096,0.03971111112170749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,3072,0.0775137808587816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,4096,0.081149333053165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,5120,0.09964622391594781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,3584,0.07211466630299886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,3072,0.03132889005872939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,2560,0.06434400214089288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,3072,0.04343822267320421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,3072,0.06618666648864746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,2048,0.05219733383920458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,2560,0.037913779417673744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,4096,0.09863822327719794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,2560,0.056735999054378934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,1536,0.04073422153790792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,2048,0.0244195560614268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,2048,0.032855110036002264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,2048,0.04919822348488701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,1536,0.019323555959595572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,1536,0.02597066594494714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,1024,0.028588443994522095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,3584,0.03571910990609063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,768,0.021873777111371357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,1536,0.04231022132767571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,1024,0.015616888801256815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,1024,0.020099555452664692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,1024,0.03421422176890903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,768,0.016180444094869826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,768,0.014030221435758801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,768,0.030088888274298772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,256,0.009751111268997192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,2560,0.02753777801990509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,512,0.013634666800498962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,512,0.02539199921819899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,256,0.012425777812798819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,128,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,256,0.02095555596881443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,51200,0.9062684377034506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,3584,0.049626668294270836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,128,0.011392888923486074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,64,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,128,0.013108444710572561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,128,0.016582222448454965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,32,0.005381333331267039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,65536,512,0.01534222231970893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,64,0.01149866647190518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,512,0.01310222182008955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,65536,32,0.015119110544522604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,256,0.011737777955002256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,65536,1.1174275080362956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,65536,0.637713803185357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,51200,0.4849795765346951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,51200,0.8791564305623373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,16384,0.28471289740668404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,16384,0.16212977303398982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,65536,0.40172799428304035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,12288,0.21424799495273164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,16384,0.11066755321290757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,12288,0.1238737768597073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,10240,0.16828978061676025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,12288,0.08115377691056994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,51200,0.32592799928453237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,10240,0.10199999809265137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,8192,0.13825778166453043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,8192,0.08646844493018256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,16384,0.24444977442423502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,7168,0.11965600649515788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,10240,0.14759022659725612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,8192,0.056072890758514404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,8192,0.12405155764685737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,7168,0.04986310998598734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,7168,0.073125335905287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,6144,0.10425066947937012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,10240,0.06901333067152235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,6144,0.04450755649142795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,6144,0.06420355372958712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,7168,0.1074062254693773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,5120,0.08851822217305501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,4096,0.07275199890136719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,12288,0.18155554930369058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,5120,0.054752889606687755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,6144,0.09683555364608765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,3584,0.06331466966205172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,4096,0.03275288807021247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,4096,0.04504088891877068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,5120,0.08253600200017293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,3072,0.05538577834765116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,4096,0.0689306656519572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,3584,0.029331554969151814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,3584,0.04039466712209914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,2560,0.04673511121008131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,3584,0.06226399872038099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,3072,0.02623555560906728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,3072,0.056403557459513344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,2560,0.02347911066479153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,51200,0.7493093278672961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,2048,0.03856622179349264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,2560,0.0492248899406857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,1536,0.0295448899269104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,5120,0.038041777080959745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,2048,0.02735911144150628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,2048,0.05113866594102648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,1536,0.02163644466135237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,2048,0.020005333754751418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,1536,0.016541333662139047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,1024,0.02285066743691762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,1536,0.03680533170700073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,768,0.01679022279050615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,65536,0.9446516036987305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,1024,0.013815999858909182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,512,0.011380444798204633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,1024,0.016202666693263583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,768,0.014716444744004143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,768,0.013478221992651621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,3072,0.03555555476082696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,512,0.013428444663683573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,512,0.011976000335481433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,768,0.026033777329656813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,512,0.02204799983236525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,128,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,2560,0.030798223283555772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,256,0.01032977799574534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,256,0.010990222295125326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,256,0.01891377733813392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,51200,128,0.01038933296998342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,128,0.009975999593734741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,32,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,128,0.014848889576064216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,64,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,51200,32,0.012738666600651212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,65536,0.3274293210771349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,65536,0.23121955659654406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,51200,0.2512302133772108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,51200,1024,0.030202666918436687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,256,0.007711999946170383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,65536,0.1402533319261339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,51200,0.17727822727627227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,16384,0.08549511432647705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,51200,64,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,16384,0.06172355678346422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,16384,0.040961778826183744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,12288,0.0651715530289544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,16384,0.061555557780795626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,12288,0.04684355523851183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,12288,0.03231911195649041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,10240,0.05592888593673706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,12288,0.047678222258885704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,8192,0.04557866520351834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,10240,0.02801688843303257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,10240,0.040123555395338274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,10240,0.04234310984611511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,65536,0.21025778187645805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,7168,0.039643555879592896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,8192,0.034536888202031456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,8192,0.023899555206298828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,6144,0.034733331865734525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,8192,0.03790400094456143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,7168,0.02125511070092519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,7168,0.030164443784289893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,51200,0.11493421925438775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,5120,0.030135999123255413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,6144,0.019881778293185763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,6144,0.026760889424218073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,6144,0.030233777231640283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,4096,0.02574666672282749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,5120,0.016960889101028442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,5120,0.023139556248982746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,5120,0.026509332987997267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,3584,0.021967111362351313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,4096,0.01953866746690538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,4096,0.015269334117571512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,4096,0.023557333482636347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,3072,0.019274666905403137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,3584,0.01499555508295695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,3584,0.017637333936161466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,2560,0.016549333930015564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,3584,0.021573333276642695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,3072,0.01311911145846049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,3072,0.0159288893143336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,2048,0.013752000199423896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,3072,0.01957155598534478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,51200,0.16668266720241973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,2560,0.012272000312805176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,1536,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,2560,0.018178666631380718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,2048,0.01132266637351778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,2048,0.01275111072593265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,1024,0.008290666672918532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,1536,0.009339555270142024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,2048,0.01698666645420922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,768,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,1536,0.010770666930410596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,1536,0.015561777684423657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,1024,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,1024,0.009176889227496253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,512,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,1024,0.014111111561457316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,768,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,768,0.008235555556085374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,256,0.0035208890007601846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,768,0.012839111189047495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,512,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,128,0.003079111170437601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,512,0.010382222632567087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,256,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,256,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,64,0.0028764444092909494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,128,0.0069013333155049225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,256,0.010326222413116032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,128,0.0064382221963670515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,16384,32,0.003136000078585413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,128,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,2560,0.014130666851997375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,64,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,65536,0.244849787818061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,32,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,7168,0.033360888560613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,51200,0.18722132841746011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,65536,0.1774995591905382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,16384,0.06404622395833333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,65536,0.12425511413150364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,51200,0.13582044177585179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,51200,0.10087200005849202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,12288,0.049471110105514526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,16384,0.048545777797698975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,16384,0.03738666574160258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,16384,512,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,10240,0.04158933295143975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,16384,0.05288889010747274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,12288,0.029663112428453233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,12288,0.03741511040263706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,8192,0.034803556071387395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,10240,0.02551288902759552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,12288,0.04261689053641426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,10240,0.03716533382733663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,8192,0.028343998723559912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,7168,0.03019555409749349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,51200,0.13878133561876085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,8192,0.021791110436121624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,65536,0.17511733373006186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,6144,0.02641600039270189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,8192,0.03315111001332601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,7168,0.01960000064637926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,7168,0.024383111132515803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,5120,0.022864888111750286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,7168,0.02953688965903388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,6144,0.018010665973027546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,4096,0.019837333096398246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,6144,0.027260444230503503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,5120,0.016338666280110676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,5120,0.0185217774576611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,3584,0.016805332567956712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,4096,0.014684443672498068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,4096,0.02139555580086178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,3584,0.014491554763582019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,3072,0.01479644411140018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,10240,0.031981332434548274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,3584,0.019621334142155118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,3072,0.012689777546458773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,3072,0.013285333083735572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,2560,0.012422222230169507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,3072,0.018584888842370775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,2560,0.011690666278203329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,6144,0.02158222264713711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,2560,0.012053333222866058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,2048,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,2560,0.017269333203633625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,1536,0.008469333251317343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,2048,0.01018488903840383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,5120,0.024215110474162634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,2048,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,4096,0.015798222687509324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,2048,0.016089777151743572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,1536,0.008798222574922774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,1536,0.009679111341635386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,768,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,1536,0.014650666051440768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,3584,0.01366222235891554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,1024,0.007746666669845581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,1024,0.007575111256705389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,512,0.004262222184075249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,1024,0.012096000214417776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,256,0.0032213332338465583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,768,0.007670222057236566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,768,0.00693866651919153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,512,0.007099555598364935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,768,0.01186222251918581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,512,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,256,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,512,0.010666666759385003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,256,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,64,0.0028604444944196274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,256,0.010129777921570672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,32,0.0028862222615215513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,128,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,128,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,128,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,64,0.006014222072230444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,65536,0.20156621932983398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,12288,32,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,1024,0.007164444360468123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,65536,0.1695502201716105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,51200,0.1565928856531779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,16384,0.054435554477903575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,65536,0.12145155005984837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,51200,0.12987732887268066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,51200,0.09680533409118652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,16384,0.03615377677811517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,16384,0.04650577902793884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,16384,0.05061777763896518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,12288,128,0.002893333426780171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,10240,0.03503733211093479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,12288,0.0280675556924608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,12288,0.035579555564456515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,8192,0.029785778787400987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,12288,0.040496001640955605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,10240,0.024306666519906785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,51200,0.12801955805884466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,65536,0.16072977913750544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,10240,0.03061244553989834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,7168,0.02552977701028188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,10240,0.035587555832333036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,8192,0.020968000094095867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,8192,0.026538666751649644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,6144,0.02241866621706221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,7168,0.019318222999572754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,8192,0.03132889005872939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,5120,0.019458666443824768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,7168,0.022980444961124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,6144,0.017557332913080852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,4096,0.016247999336984422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,6144,0.020607110526826646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,7168,0.029069334268569946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,5120,0.01572355628013611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,6144,0.026307556364271376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,5120,0.017755554782019723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,12288,0.04229422079192268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,5120,0.02323377794689602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,4096,0.015375110838148328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,3072,0.012302222351233164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,4096,0.02097333305411869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,3584,0.014434667097197639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,3584,0.0196079992585712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,3072,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,3072,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,2560,0.010339555641015371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,3072,0.018345778187115986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,2560,0.01130222198035982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,2560,0.01180622229973475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,2560,0.017147555947303772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,2048,0.010652444428867765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,2048,0.009330666727489894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,2048,0.015801777442296345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,3584,0.013843555417325763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,1536,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,1024,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,1536,0.008358221915033128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,3584,0.013438222308953604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,1536,0.014184888866212634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,768,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,1024,0.007682667010360294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,1024,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,1024,0.011852444046073489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,768,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,768,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,2048,0.008805333740181392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,768,0.011384888655609555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,512,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,512,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,512,0.01034311122364468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,1536,0.007824888659848107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,128,0.002806222273243798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,256,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,256,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,256,0.00963466697269016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,64,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,4096,0.014306666122542487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,128,0.00674222202764617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,32,0.0028942223224374983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,128,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,128,0.008307555483447181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,64,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,10240,32,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,65536,0.17946222093370226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,512,0.003539555602603488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,51200,0.14380978213416204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,65536,0.12626666492886013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,16384,0.04821599854363335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,65536,0.07243911425272624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,51200,0.0966942244105869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,10240,256,0.0031662223239739737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,12288,0.03752355443106757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,51200,0.058127111858791776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,16384,0.023174222972657945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,16384,0.03593066665861342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,10240,0.03147466646300422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,12288,0.01872177753183577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,12288,0.027894222074084814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,16384,0.046763555871115796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,8192,0.026661333110597398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,12288,0.037979556454552546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,10240,0.023835556374655828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,10240,0.016515556308958266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,51200,0.11507466104295518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,65536,0.14433599842919245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,7168,0.02295200030008952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,10240,0.033744888173209295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,6144,0.020111999577946134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,8192,0.014711111783981323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,7168,0.013516444298956128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,8192,0.029355555772781372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,5120,0.01703288820054796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,7168,0.018212444252438016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,6144,0.012523555921183692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,6144,0.016527111331621807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,7168,0.02697333362367418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,4096,0.01477599971824222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,5120,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,6144,0.0244835548930698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,3584,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,4096,0.010327999790509542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,4096,0.01275288893116845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,5120,0.022441777918073866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,3072,0.010646222366227044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,4096,0.02013777858681149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,3584,0.011783999701340994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,2560,0.009282666775915358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,3584,0.01871555546919505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,3072,0.009006222089131674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,3072,0.011170667078759937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,2048,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,2560,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,2560,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,2560,0.01651822196112739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,8192,0.02051999999417199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,2048,0.00925244473748737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,1536,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,2048,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,2048,0.015160888433456421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,1536,0.007010666860474481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,1536,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,1536,0.012732444538010491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,1024,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,3584,0.009763555394278632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,768,0.0042204442951414315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,1024,0.011867555479208628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,768,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,768,0.006034666465388403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,3072,0.01754755609565311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,512,0.003208000005947219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,768,0.01070577816830741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,256,0.0029022221763928733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,512,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,512,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,512,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,128,0.002789333255754577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,256,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,256,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,1024,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,64,0.002536888958679305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,128,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,8192,32,0.002830222249031067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,128,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,5120,0.01439377831088172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,128,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,64,0.005663110978073544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,1024,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,8192,32,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,51200,0.1256213320626153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,65536,0.12430132759941949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,65536,0.068904889954461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,16384,0.04451377855406868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,51200,0.09582044680913289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,51200,0.05626577801174588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,16384,0.03495288888613383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,16384,0.021909332937664453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,12288,0.03494044476085239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,16384,0.04550488789876302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,10240,0.02880355715751648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,51200,0.11436088879903157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,12288,0.01811911165714264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,8192,0.024780443973011438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,10240,0.015909334023793537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,12288,0.037603557109832764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,10240,0.03340533375740051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,7168,0.021480000681347314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,8192,0.01998488936159346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,65536,0.16113689210679796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,8192,0.014151111245155334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,6144,0.018699554933442008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,8192,0.02881511052449544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,7168,0.013096888860066732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,7168,0.01753155555990007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,5120,0.01588711142539978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,7168,0.02647288971477085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,6144,0.012126222252845764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,6144,0.023804444405767653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,4096,0.013507555756303998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,5120,0.01129155523247189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,5120,0.014077333940400017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,5120,0.021627555290857952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,65536,0.14281866285536024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,3584,0.011056889262464313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,4096,0.009957333405812582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,4096,0.012792000340090858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,12288,0.026867555247412786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,3072,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,4096,0.019838222199016146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,10240,0.023238221804300945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,3584,0.009007111191749573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,2560,0.008760000268618265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,3584,0.018425777554512024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,3072,0.011039999624093374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,2048,0.007595555649863348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,2560,0.009936889012654623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,2560,0.008018666671382057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,2560,0.01612711118327247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,1536,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,2048,0.007419555551475949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,2048,0.008485332959228093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,6144,0.016031111280123394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,1024,0.004580444345871608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,2048,0.014861333701345654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,1536,0.0069822221994400024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,1536,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,768,0.003937777959638172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,1536,0.012485333614879184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,1024,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,1024,0.006478222293986215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,512,0.0035048888789282907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,768,0.006346666564544042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,768,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,768,0.01071111112833023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,3584,0.011772444678677453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,256,0.0031742221779293488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,512,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,3072,0.00833511104186376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,512,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,512,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,3072,0.01715555621518029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,256,0.006135111053784688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,256,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,64,0.002537777854336633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,256,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,128,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,32,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,128,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,65536,0.14477154943678114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,64,0.005648889061477449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,32,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,65536,0.11698755953047012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,51200,0.11434311336941189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,1024,0.011410666836632622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,65536,0.06649777624342176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,16384,0.03546844588385688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,51200,0.08945155806011623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,51200,0.0543431109852261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,7168,128,0.00282577777074443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,16384,0.03391111228201125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,12288,0.02804533309406704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,16384,0.020997333857748244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,65536,0.14182044400109187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,10240,0.023296000228987798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,51200,0.11354400051964654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,7168,128,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,12288,0.017362667454613578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,16384,0.045351111226611666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,12288,0.02561155623859829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,8192,0.019966221517986722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,12288,0.03681333197487725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,10240,0.015447111593352424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,7168,0.01722311145729489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,10240,0.021993777818149988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,8192,0.013772444592581855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,8192,0.018896889355447557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,6144,0.014926221635606555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,8192,0.02822222312291463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,7168,0.012776888906955719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,7168,0.01699911057949066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,5120,0.013006221916940479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,7168,0.025573333104451496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,6144,0.011892444557613797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,6144,0.015666666958067153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,4096,0.011380444798204633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,6144,0.023521777656343248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,5120,0.010724444356229572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,5120,0.013911111487282647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,5120,0.021016000045670405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,4096,0.009335111412737105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,4096,0.01240888900227017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,3072,0.008989333278603023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,4096,0.019560888409614563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,3584,0.00865955568022198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,3584,0.011517333487669626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,2560,0.007752888732486301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,3584,0.01789155602455139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,3072,0.008303999900817871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,3072,0.010709332923094431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,2048,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,2560,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,2560,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,1536,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,2560,0.015487111277050443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,2048,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,2048,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,10240,0.03268977668550279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,1024,0.003960000144110786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,2048,0.01422044469250573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,1536,0.006896889044178857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,768,0.0038177776667806837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,1536,0.007255111303594377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,1024,0.006559111177921295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,1536,0.012375999655988483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,1024,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,512,0.003173333282272021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,1024,0.011056000159846412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,768,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,3584,0.010073777702119615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,256,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,768,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,512,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,768,0.01074399964676963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,128,0.002528000002106031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,256,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,512,0.010416000253624385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,256,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,64,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,6144,32,0.00249422217408816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,256,0.009883555273214975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,128,0.0058746664888328975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,128,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,128,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,3072,0.017172444197866652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,64,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,32,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,51200,0.09607822365230984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,65536,0.06422577963935004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,51200,0.0866968896653917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,51200,0.052128887838787503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,16384,0.03735555542839898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,65536,0.14126043849521214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,51200,0.11347911092970107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,6144,512,0.0057297779454125305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,16384,0.02052533295419481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,16384,0.03193599979082743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,12288,0.029663999875386555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,10240,0.021272889441914026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,16384,0.045129779312345714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,12288,0.016833777228991192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,12288,0.02436888880199856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,65536,0.11984266175164117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,12288,0.03635466761059231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,8192,0.01816444430086348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,10240,0.015145777000321282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,10240,0.02146133283774058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,7168,0.015463999576038785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,10240,0.032199111249711775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,65536,0.1121893326441447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,6144,0.013776889277829064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,8192,0.013302221894264221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,8192,0.01850755512714386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,7168,0.012390221986505719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,7168,0.01679022279050615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,8192,0.02752977775202857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,5120,0.011968889170222811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,7168,0.0251857770813836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,4096,0.010465777582592435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,6144,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,5120,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,6144,0.023307555251651343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,5120,0.013573333621025085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,3584,0.008965333302815756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,5120,0.02095555596881443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,4096,0.012089778151777057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,4096,0.008647110727098253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,4096,0.019462222854296368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,3584,0.008634666601816813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,3584,0.011374221907721626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,2560,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,3584,0.01755644381046295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,3072,0.007954667011896769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,3072,0.010049777726332346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,2048,0.006360000206364526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,3072,0.01684977776474423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,2560,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,2560,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,2560,0.01575555569595761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,2048,0.00720977783203125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,2048,0.007676444119877285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,1024,0.0038622220357259116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,2048,0.013788444300492605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,1536,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,1536,0.0070266665683852295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,768,0.003536000019974179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,6144,0.01514400045077006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,1536,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,1024,0.0063733334342638654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,1024,0.006413333531883027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,768,0.006230222268237009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,1024,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,768,0.005993777679072485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,256,0.0028515555378463534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,768,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,3072,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,512,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,512,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,128,0.0025031111306614345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,256,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,512,0.009671111073758867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,256,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,64,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,32,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,256,0.009019555317031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,128,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,128,0.005454222361246745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,1536,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,65536,0.0945697757932875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,64,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,5120,32,0.00573955559068256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,51200,0.07647022273805407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,65536,0.07087466451856825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,65536,0.06185688575108846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,16384,0.030280888080596924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,51200,0.05478577812512716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,51200,0.04996711015701294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,5120,512,0.0031724443866146933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,12288,0.023974221613672044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,16384,0.021313778228229944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,16384,0.019692444139056735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,10240,0.017981333865059745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,51200,0.11281243960062663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,16384,0.04416977696948581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,12288,0.01628977722591824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,12288,0.016767111089494493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,8192,0.015440889530711703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,10240,0.01463733282354143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,10240,0.01481244464715322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,10240,0.03101955519782172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,7168,0.013201778133710226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,8192,0.013510222236315409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,128,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,8192,0.027112889620992873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,6144,0.011688888900809817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,7168,0.011776888536082374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,7168,0.012309333516491784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,7168,0.025077333052953083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,5120,0.010613333019945355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,6144,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,65536,0.14153867297702366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,5120,0.010216889282067617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,5120,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,4096,0.0092604441775216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,5120,0.020601777566803824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,12288,0.03554844525125291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,3584,0.008088889221350351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,4096,0.00907733374171787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,4096,0.018760000665982563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,8192,0.012917333179050021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,3072,0.007559110720952352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,3584,0.0083155557513237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,3584,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,2560,0.006561777657932705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,3072,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,3072,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,3072,0.016093333562215168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,6144,0.011340444286664328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,2048,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,2560,0.007375111182530721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,2560,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,6144,0.023063111636373732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,1536,0.00464088883664873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,2560,0.014464888307783337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,2048,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,2048,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,1024,0.003504000190231535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,1536,0.006682666639486949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,2048,0.013335111240545908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,768,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,1536,0.006331555545330048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,1536,0.01220355596807268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,1024,0.006260444306664997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,4096,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,512,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,1024,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,768,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,1024,0.010871999793582492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,768,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,512,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,768,0.010777778095669217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,128,0.002502222235004107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,512,0.005383111122581694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,512,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,256,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,256,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,3584,0.017998221847746108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,256,0.0092257774538464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,128,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,128,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,65536,0.08322755495707194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,64,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,4096,32,0.00536088893810908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,51200,0.06712266471650866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,65536,0.0679253339767456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,65536,0.06093688805898031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,51200,0.049119111564424306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,65536,0.14104621940188938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,51200,0.05285955468813578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,16384,0.028608888387680054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,256,0.002464888824356927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,12288,0.02292888859907786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,51200,0.11327021651797825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,16384,0.01947377787695991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,16384,0.02029511166943444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,10240,0.019881778293185763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,64,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,16384,0.043706665436426796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,4096,32,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,12288,0.016164445214801364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,12288,0.016131555040677387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,8192,0.014131555954615274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,12288,0.035409778356552124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,7168,0.012819555898507437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,128,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,10240,0.01458044515715705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,8192,0.0129457778400845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,6144,0.011340444286664328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,10240,0.030978666411505804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,8192,0.012858666479587555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,7168,0.011367999845080905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,8192,0.026551999979548987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,5120,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,7168,0.012059555285506778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,6144,0.01033333357837465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,7168,0.024697777297761705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,4096,0.008967111508051554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,6144,0.011286222272449069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,6144,0.022295110755496558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,5120,0.009775111244784461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,3584,0.00775911079512702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,5120,0.02087822225358751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,4096,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,4096,0.018858666221300762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,3072,0.007091555330488417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,3584,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,3584,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,3584,0.017197334104114108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,2560,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,3072,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,3072,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,10240,0.014442667365074158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,2048,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,3072,0.015504888362354703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,2560,0.007340444458855524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,2560,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,1536,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,2560,0.014163555370436775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,2048,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,2048,0.00645955569214291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,1024,0.0035182221068276297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,2048,0.013009777499569787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,1536,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,5120,0.009309333231714036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,1536,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,768,0.003089777711364958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,1536,0.011736888852384357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,1024,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,4096,0.00869599978129069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,1024,0.005718222094906702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,512,0.0028328889360030494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,768,0.006321777900060018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,768,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,256,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,512,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,768,0.010672888822025724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,512,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,512,0.009678222239017487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,256,0.005682666682534748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,256,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,64,0.0021955555097924336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,256,0.009474666582213508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,128,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,128,0.004949333353175057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,32,0.002542222125662698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,128,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,64,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3584,32,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,65536,0.06603378057479858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,51200,0.061979558732774526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,65536,0.05978666411505806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,16384,0.023740443918440077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,65536,0.14083022541469997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,51200,0.05105333195792305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,51200,0.047641777329974704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,1024,0.01108977778090371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,12288,0.019751111666361492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,16384,0.019800000720553927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,51200,0.11301955911848281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3584,128,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,10240,0.0162808895111084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,12288,0.016050666570663452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,16384,0.043410665459103055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,8192,0.014484445254007975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,12288,0.034663110971450806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,10240,0.014415999253590902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,10240,0.014160000615649752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,7168,0.012517333030700684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,65536,0.07589688565995958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,10240,0.030997332599427965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,8192,0.012072888513406118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,8192,0.01256977766752243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,7168,0.011078221930397881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,8192,0.026513778501086768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,7168,0.011439999772442712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,5120,0.009722666607962714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,7168,0.024433778391944036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,6144,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,6144,0.010695111420419482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,4096,0.008549333446555668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,5120,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,5120,0.00961511085430781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,5120,0.020227554771635268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,16384,0.019288000133302476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,3584,0.007576000359323289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,4096,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,4096,0.008359111017651027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,3072,0.00702400008837382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,4096,0.01795288920402527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,12288,0.015834665960735746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,3584,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,2560,0.00620266670982043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,3584,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,3584,0.01642933322323693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,3072,0.007663999994595845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,3072,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,2048,0.005379555539952383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,3072,0.015365334020720588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,2560,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,2560,0.014365333649847241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,2048,0.00703111125363244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,1536,0.004578666554556953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,2048,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,2048,0.012888000243239932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,1024,0.0038213332494099936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,1536,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,6144,0.02288355595535702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,1536,0.006225777582989798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,1536,0.011878222227096558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,1024,0.0063822223908371395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,1024,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,1024,0.011047110789351992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,512,0.0029120000286234748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,768,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,768,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,512,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,256,0.002540444541308615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,512,0.0053102220926019884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,512,0.009681777821646797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,2560,0.007443555527263218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,128,0.002505777817633417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,256,0.009709333380063375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,128,0.004954666727119022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,64,0.0022524444179402455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,128,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,128,0.007585778004593319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,768,0.0034088889757792154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,65536,0.06364888615078397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,64,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,32,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,51200,0.05178222060203552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,65536,0.06302577919430204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,65536,0.05840444564819336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,6144,0.01110400011142095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,51200,0.04917066627078586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,51200,0.04652888907326592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,65536,0.14014932844373915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,3072,256,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,12288,0.017958222164048087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,16384,0.018992889258596633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,51200,0.11259733306037055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,16384,0.01905155513021681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,10240,0.015434667468070984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,16384,0.04274310999446445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,3072,32,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,12288,0.015449777245521545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,12288,0.015680889288584392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,8192,0.013549333645237817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,10240,0.014075555735164218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,12288,0.03438933359252082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,10240,0.013767110804716745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,7168,0.01221244451072481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,8192,0.012390221986505719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,10240,0.03063911199569702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,8192,0.011148444480366178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,6144,0.011117333339320289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,8192,0.02629866699377696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,7168,0.010819555156760745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,768,0.01074399964676963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,5120,0.00962400022480223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,7168,0.01107377807299296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,6144,0.00980977796845966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,7168,0.024251555403073628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,4096,0.008316444853941599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,5120,0.00904799997806549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,6144,0.022260444031821355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,16384,0.02090488870938619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,3584,0.007521777517265744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,5120,0.009186666872766282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,5120,0.01957155598534478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,4096,0.008358221915033128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,3072,0.006695111178689533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,4096,0.008077333370844523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,3584,0.00795555528667238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,4096,0.017868444323539734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,2560,0.00600355573826366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,3584,0.007770666645632849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,3072,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,3584,0.01620533400111728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,2048,0.005016889009210798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,3072,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,3072,0.015167999598715039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,2560,0.007592000067234039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,2560,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,1536,0.00424177779091729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,2048,0.007224889265166388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,2560,0.014496889379289416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,1024,0.003536000019974179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,2048,0.013056000073750814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,1536,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,768,0.0031582222630580268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,1536,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,1024,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,1024,0.005639111002286275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,1024,0.010759111079904767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,6144,0.010213333699438307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,512,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,768,0.006223111102978389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,768,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,256,0.0024871110088295406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,768,0.010668444136778513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,512,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,512,0.005484444399674733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,512,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,128,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,256,0.00574222207069397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,256,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,64,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,256,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,128,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2560,32,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,128,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,65536,0.052424887816111244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,128,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,2048,0.006400888992680444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,64,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,1536,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,51200,0.04171466827392578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,65536,0.05707377857632107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,65536,0.06085777944988675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,16384,0.01735733283890618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,51200,0.04548622171084086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,51200,0.047262221574783325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,65536,0.14018932978312174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,12288,0.013526221944226159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,51200,0.1123724381128947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,16384,0.018479110466109384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,16384,0.018800000349680584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,10240,0.01259200026591619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,16384,0.04273600048489041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,12288,0.014656888114081489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,8192,0.012858666479587555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,12288,0.015261333849694995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,12288,0.034663110971450806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,10240,0.012768000364303589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,10240,0.013439999686347114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,7168,0.011538666983445486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,10240,0.030264000097910564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,8192,0.011411555939250521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,8192,0.011804444094498953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,6144,0.010318222145239512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,7168,0.010381333529949188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,8192,0.02609599961174859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,7168,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,5120,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,7168,0.024527110987239416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,6144,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2560,32,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,6144,0.009659555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,4096,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,6144,0.021755556265513103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,5120,0.009030222064918941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,5120,0.008656889200210571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,4096,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,5120,0.019201777047581144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,3072,0.006227555374304454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,4096,0.007954667011896769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,4096,0.017224889662530687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,2560,0.005346666607591841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,3584,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,3584,0.015875554747051664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,3072,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,2048,0.004568888909286923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,3072,0.015288000305493673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,2560,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,2560,0.0068959999415609576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,1536,0.003882666842805015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,2560,0.014115555418862237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,2048,0.006367110957702001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,2048,0.013395555317401886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,1536,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,1024,0.003169777699642711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,1536,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,1024,0.006308444258239534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,3584,0.007156444092591603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,768,0.002967111145456632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,1024,0.011412444214026133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,3584,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,768,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,512,0.0025271111064487034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,768,0.00553955551650789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,768,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,256,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,512,0.005958222266700532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,512,0.0052826665341854095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,512,0.009661333428488838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,128,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,256,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,2048,0.007110222346252865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,256,0.009343110852771336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,1536,0.006168889088763132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,128,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,128,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,32,0.002161777681774563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,64,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,1024,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,65536,0.0421013335386912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,2048,32,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,65536,0.05748799774381849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,3072,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,51200,0.03355733222431607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,65536,0.05570222271813286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,65536,0.14052089055379233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,16384,0.013953778478834363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,51200,0.044774221049414746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,51200,0.04560977882809109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,256,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,12288,0.0123831108212471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,16384,0.01721955504682329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,51200,0.1121022171444363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,2048,64,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,16384,0.01848355597919888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,10240,0.011226666470368704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,16384,0.04246133234765795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,12288,0.013735111388895245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,12288,0.014873777826627096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,8192,0.009763555394278632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,10240,0.012883555557992725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,10240,0.012464000119103326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,7168,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,8192,0.011024889018800525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,8192,0.011035555766688453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,8192,0.025754666990704004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,6144,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,7168,0.01036266651418474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,7168,0.010063110954231685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,5120,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,6144,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,6144,0.009353777600659264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,6144,0.02108088963561588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,4096,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,5120,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,5120,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,128,0.005636444522274866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,5120,0.018904000520706177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,4096,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,3584,0.006498666687144174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,4096,0.007755555212497711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,3072,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,4096,0.017484444710943434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,3584,0.008017777568764156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,3584,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,12288,0.034266667233573064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,2560,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,3584,0.015832889411184523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,3072,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,2048,0.004551110996140374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,3072,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,2560,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,3072,0.014926221635606555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,1536,0.004017777740955353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,2560,0.006801777829726537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,2560,0.013838222457302941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,2048,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,2048,0.006319111006127463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,1024,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,2048,0.013009777499569787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,7168,0.02366666661368476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,768,0.003173333282272021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,1536,0.011766222616036734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,1024,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,1024,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,1024,0.0107351111041175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,768,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,768,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,768,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,512,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,256,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,512,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,10240,0.030121776792738173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,128,0.0021262222694026097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,256,0.005048000150256687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,256,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,256,0.008957333034939235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,64,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,128,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,32,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,128,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,128,0.007690666450394525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,65536,0.029025776518715754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,1536,0.00666133314371109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,1536,0.006078222145636876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,51200,0.023019555542204116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,32,0.004956444518433677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,65536,0.05468800001674228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1536,512,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,16384,0.012100444071822695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,65536,0.055382221937179565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,51200,0.044403556320402354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,512,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,12288,0.012524444195959302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,51200,0.1116151147418552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,16384,0.01631733278433482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,16384,0.04247644543647766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,12288,0.01444355481200748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,10240,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,12288,0.033294220765431724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1536,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,10240,0.012083555261294046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,8192,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,10240,0.029647111892700195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,10240,0.0123848890264829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,8192,0.010847110715177325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,7168,0.009480888644854227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,51200,0.04348177711168925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,8192,0.02518400053183238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,16384,0.017830222845077515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,6144,0.008608000146018134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,7168,0.010068444742096795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,7168,0.00998755544424057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,7168,0.023381332556406658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,5120,0.007916444705592262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,6144,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,12288,0.013408888545301227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,6144,0.009295110901196798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,4096,0.00702400008837382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,6144,0.021176889538764954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,5120,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,5120,0.018595554762416415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,4096,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,4096,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,4096,0.01684266659948561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,8192,0.010575999816258749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,3584,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,3072,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,65536,0.140611555841234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,3584,0.016349333855840895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,2560,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,3072,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,3072,0.007076444725195567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,3072,0.014767110347747803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,2560,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,2560,0.014126222994592456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,5120,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,1536,0.00388355553150177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,2048,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,2048,0.012772444221708508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,3584,0.006230222268237009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,1024,0.0032142222755485107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,1536,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,1536,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,1536,0.01204711116022534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,768,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,3584,0.008010666403505538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,1024,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,1024,0.010713777608341642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,768,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,768,0.005502222312821283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,768,0.010209778116808997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,2048,0.004505777938498391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,256,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,2560,0.006667555620272954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,512,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,512,0.010028444230556488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,128,0.0021235555824306277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,256,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,256,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,256,0.009736888938479954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,64,0.0018275556051068837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,128,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,32,0.001863111017478837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,128,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,65536,0.025391111771265667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,64,0.004967111266321606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,1024,0.006302222195598815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,32,0.005041777673694822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,51200,0.0210506667693456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,65536,0.05436977744102478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,65536,0.05342222253481547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,65536,0.1399866607454088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,51200,0.04400444361898634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,51200,0.04281333420011732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,16384,0.012368000215954252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,512,0.005401777724424998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,12288,0.010678222609890832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,16384,0.016161777906947665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,2048,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,16384,0.016717332932684157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,10240,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,16384,0.04148888919088576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,12288,0.013413333230548434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,1024,128,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,8192,0.008839999636014303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,10240,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,10240,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,10240,0.028953777419196233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,8192,0.011049778097205691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,7168,0.0083155557513237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,8192,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,8192,0.024717332588301763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,1024,512,0.002507555608948072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,6144,0.0075351107451650835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,7168,0.010370666782061258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,7168,0.009977777798970541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,7168,0.02273511058754391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,5120,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,6144,0.009697777529557547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,51200,0.11138044463263617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,5120,0.008574222524960836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,5120,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,4096,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,5120,0.018900444110234577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,12288,0.03360355562633938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,3584,0.0064071110553211635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,4096,0.00777688870827357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,4096,0.016713778177897137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,3584,0.007962666451931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,3584,0.0075253330998950535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,3072,0.005542222410440445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,3584,0.016197333733240765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,3072,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,3072,0.007186666958861881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,6144,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,3072,0.014780445231331719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,6144,0.021077333225144282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,2048,0.0042453333735466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,2560,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,2560,0.006654222392373615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,12288,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,2560,0.013773333695199756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,1536,0.0038328886859946777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,2048,0.0070053330726093715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,2048,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,1024,0.003157333367400699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,4096,0.00833777752187517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,2048,0.0127368892232577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,1536,0.006672888994216919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,768,0.0028693332440323303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,1536,0.01147377739350001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,1024,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,1024,0.005670222143332164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,512,0.002472888885272874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,1024,0.01075111081202825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,768,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,768,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,768,0.010316444767846001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,512,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,2560,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,128,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,512,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,256,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,256,0.0057502223385704895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,64,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,256,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,128,0.005639111002286275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,128,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,128,0.00758755538198683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,64,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,32,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,51200,0.015954666667514376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,768,1536,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,65536,0.05287733342912462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,65536,0.05407022105322944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,16384,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,65536,0.13902578088972303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,51200,0.040469331873787776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,51200,0.04256177941958109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,768,256,0.002193777718477779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,12288,0.00870666652917862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,16384,0.01651822196112739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,51200,0.11039200094011094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,512,0.005643555687533484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,10240,0.00812266684240765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,16384,0.041251556740866765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,12288,0.013434666726324292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,12288,0.013451555536852943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,8192,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,10240,0.012096888489193387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,10240,0.0284862220287323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,7168,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,65536,0.019319110446506076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,8192,0.01073155552148819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,8192,0.010440000229411654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,6144,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,8192,0.02456888887617323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,7168,0.010327111515733931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,7168,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,5120,0.006637333167923822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,6144,0.009588444398509132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,7168,0.022611555125978258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,6144,0.009306666751702627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,4096,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,6144,0.020631111330456205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,5120,0.008973333570692275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,5120,0.008399999803966945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,3584,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,4096,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,16384,0.01609244445959727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,4096,0.007878222399287755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,3072,0.005542222410440445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,4096,0.016809778081046212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,3584,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,3584,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,10240,0.012091555529170565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,3584,0.015806222955385845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,2560,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,3072,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,3072,0.015115555789735583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,2048,0.004200889004601372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,2560,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,2560,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,2560,0.013794666363133324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,1536,0.0038684445122877755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,2048,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,2048,0.0063484443558586975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,2048,0.012412444584899478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,1024,0.0031964443624019623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,1536,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,1536,0.011750222080283694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,5120,0.018583999739752877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,768,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,1024,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,1024,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,1024,0.0107004443804423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,512,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,768,0.006132444573773279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,12288,0.03348088926739163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,3072,0.007032000356250339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,768,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,256,0.0021102221475707158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,768,0.010049777726332346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,512,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,128,0.002003555496533712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,512,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,256,0.005025777965784073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,256,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,64,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,256,0.009649777577983009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,512,32,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,128,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,128,0.004978666702906291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,65536,0.0151573336786694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,128,0.007339555356237624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,64,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,51200,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,512,32,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,65536,0.047055999437967934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,16384,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,65536,0.052008890443378024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,65536,0.1388346619076199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,51200,0.040241777896881104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,12288,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,16384,0.016303999556435477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,16384,0.015853333804342482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,51200,0.11037955681482951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,16384,0.04131644301944309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,12288,0.013175999952687157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,12288,0.013447999954223633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,12288,0.0328435566690233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,10240,0.012100444071822695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,10240,0.012069332930776807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,10240,0.02875466810332404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,8192,0.010398222340477837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,7168,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,1536,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,8192,0.024708444873491924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,7168,0.010045333041085137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,51200,0.03836089041497972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,6144,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,7168,0.009881777895821465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,7168,0.023039110832744177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,5120,0.006507555643717448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,6144,0.009679999616410997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,10240,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,6144,0.009115555220180089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,4096,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,6144,0.020590222544140287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,5120,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,5120,0.008630221916569604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,3584,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,512,0.009478222164842818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,5120,0.01924088928434584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,4096,0.00814222213294771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,4096,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,3072,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,4096,0.016875555117925007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,3584,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,2560,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,3584,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,8192,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,3072,0.007583111524581909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,3584,0.015744000673294067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,3072,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,2048,0.0047173334492577445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,3072,0.014597333139843412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,2560,0.007242666350470648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,1536,0.0037475555307335327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,2560,0.006682666639486949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,2048,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,1024,0.003157333367400699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,2560,0.014139556222491793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,2048,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,1536,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,2048,0.012772444221708508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,768,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,1536,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,1536,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,1024,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,1024,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,512,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,768,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,768,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,768,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,256,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,512,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,512,0.010021333065297868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,128,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,256,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,256,0.004863111095296012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,256,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,64,0.0017928888814316855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,128,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,128,0.004953777624501122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,32,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,65536,0.01167555567291048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,128,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,64,0.005026666654480828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,32,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,51200,0.010701333483060202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,65536,0.048947556151284106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,65536,0.04688088761435615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,65536,0.13804622491200766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,16384,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,51200,0.037984887758890785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,51200,0.04029066695107354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,256,8192,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,51200,0.10962844557232326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,12288,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,16384,0.01588711142539978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,16384,0.01613066593805949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,10240,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,16384,0.040977779361936785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,12288,0.013432000246312885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,12288,0.013395555317401886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,8192,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,12288,0.03261244297027588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,10240,0.011734222372372946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,10240,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,7168,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,10240,0.029096888171301946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,8192,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,8192,0.010283555421564314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,6144,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,8192,0.02438933319515652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,7168,0.010051555931568146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,7168,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,5120,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,7168,0.023016888234350417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,6144,0.009378666679064432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,4096,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,6144,0.008973333570692275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,1024,0.0107342220014996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,5120,0.008986666798591614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,6144,0.020597333709398907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,5120,0.008567999634477828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,256,512,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,4096,0.00810933361450831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,3072,0.004595555365085602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,5120,0.01928444372283088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,4096,0.016697777642144095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,3584,0.007698666718271043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,2560,0.004953777624501122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,3584,0.016179554992251925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,3072,0.0075902218619982404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,3072,0.0069066666894488865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,2048,0.0047084444926844705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,3072,0.01510933372709486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,2560,0.007154666715198093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,2560,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,1536,0.003528000166018804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,2048,0.0069022224181228215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,2048,0.006143110907740063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,2048,0.013415111435784234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,1024,0.002982222164670626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,1536,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,1536,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,1536,0.011359111302428775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,1024,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,768,0.0028044444819291434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,1024,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,3584,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,512,0.0024702221983008915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,1024,0.010689777632554373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,4096,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,768,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,768,0.00527555536892679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,256,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,768,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,3584,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,512,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,128,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,512,0.0053671110007498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,256,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,512,0.00943022221326828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,64,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,256,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,128,32,0.0018151110659042995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,256,0.009669333696365356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,128,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,128,0.004874666531880696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,65536,0.00829777783817715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,51200,0.007958222594526079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,16384,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,128,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,2560,0.014122666584120857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,12288,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,64,0.004741333425045013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,10240,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,128,32,0.004558222161398994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,8192,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,7168,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,65536,0.04861066738764445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,51200,0.0397093329164717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,6144,0.005179555465777715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,16384,0.015804444750150044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,5120,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,12288,0.013456888496875763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,10240,0.011701333026091257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,4096,0.004554666578769684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,3584,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,3072,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,7168,0.010000000397364298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,2560,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,5120,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,2048,0.004522666749027041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,4096,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,1536,0.0035137778355015647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,3072,0.0068808889223469635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,1024,0.0028888889484935333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,2560,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,768,0.0025235555238193935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,2048,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,512,0.002492444382773505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,1536,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,256,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,1024,0.005325333525737126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,128,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,768,0.0053680001033677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,64,0.0018151110659042995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,512,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,64,32,0.001790222194459703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,256,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,65536,0.011448000040319232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,128,0.004599110947714912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,51200,0.00998488896422916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,64,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,16384,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,8192,0.010080888867378235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,12288,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,65536,0.048992888795004956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,6144,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,10240,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,51200,0.03912622067663405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,8192,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,3584,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,7168,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,16384,0.016182222300105624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,6144,0.004963555683692296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,5120,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,12288,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,4096,0.004552000098758274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,10240,0.012062222593360476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,8192,0.0101724440852801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,3584,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,7168,0.00941422250535753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,6144,0.00925155563486947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,5120,0.008344888687133789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,2048,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,4096,0.007342221836249034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,3584,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,1536,0.003559999995761447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,3072,0.006706666615274217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,1024,0.0028764444092909494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,768,0.0025315555847353404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,2560,0.006562666760550604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,512,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,2048,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,256,0.0019964445382356644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,1024,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,128,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,768,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,64,32,0.004713777866628435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,64,0.001810666587617662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,512,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,32,0.001826666709449556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,256,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,128,0.004625777817434735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,64,0.004861333303981357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,32,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,3072,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,128,32,2560,0.004556444370084339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,16384,0.3719644546508789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,16384,0.20322933461931017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,51200,0.6118808852301704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,12288,0.2805946667989095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,128,32,1536,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,16384,0.13459022839864096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,51200,1.154211574130588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,10240,0.23637066947089302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,12288,0.10046755605273777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,10240,0.1273217731051975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,10240,0.08447821935017903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,8192,0.19400266806284586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,51200,0.3979182243347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,7168,0.16958133379618326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,8192,0.1076800028483073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,8192,0.06856089168124728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,7168,0.09191289212968613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,6144,0.14621244536505804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,7168,0.0617777771419949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,5120,0.12221066157023112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,6144,0.05346222387419807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,6144,0.07920355266994901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,5120,0.04699644446372986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,5120,0.06678133540683322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,12288,0.15034133858150905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,3584,0.08768266439437866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,4096,0.05537866552670797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,4096,0.04317333300908407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,3072,0.07623733414544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,2560,0.06373244524002075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,3584,0.05069244570202298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,3072,0.04305155409706963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,2048,0.05149866806136238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,2560,0.027296889159414504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,3072,0.03052622079849243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,2560,0.03738755649990506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,1536,0.04061511158943176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,1024,0.02829155657026503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,2048,0.03240888979699876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,1536,0.01867555578549703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,2048,0.02330222229162852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,768,0.02165066699186961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,1536,0.025560888979170058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,512,0.015048889650238885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,1024,0.015695111619101632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,256,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,1024,0.019729778170585632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,4096,0.09952444500393337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,768,0.016565332810084026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,512,0.012917333179050021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,128,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,256,0.011310222248236338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,64,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,65536,32,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,128,0.011353777514563667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,128,0.011415111521879831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,3584,0.034867554903030396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,64,0.011415111521879831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,32,0.015209777487648858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,65536,768,0.01387733303838306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,65536,0.6315226554870605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,51200,0.48912800682915586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,51200,0.8779635429382324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,256,0.012423999607563019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,16384,0.28413600391811794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,16384,0.1633626619974772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,12288,0.21487289004855684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,16384,0.10638755559921265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,65536,0.4100880093044705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,10240,0.16714044411977133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,12288,0.11966933144463433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,12288,0.08173866404427423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,8192,0.13762222396002874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,51200,0.32317156261867946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,10240,0.07664444711473253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,8192,0.08630222082138062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,8192,0.055810663435194224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,10240,0.104403555393219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,7168,0.1192577812406752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,65536,512,0.013463999662134381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,6144,0.10272000233332317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,5120,0.08765688869688247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,7168,0.050717333952585854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,7168,0.07475466860665216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,6144,0.063701331615448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,6144,0.043699555926852755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,4096,0.07299288776185778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,5120,0.037744889656702675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,5120,0.0536097784837087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,4096,0.03226577904489305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,3072,0.05516622132725186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,4096,0.04475111100408766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,2560,0.04637688729498121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,3584,0.028929776615566675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,3072,0.025828444295459326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,2048,0.03787022166781955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,3072,0.03504622313711379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,2560,0.03102933367093404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,1536,0.029976000388463337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,2048,0.01979377700222863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,2048,0.027127999398443434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,1024,0.022495999932289124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,2560,0.023051554958025616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,1536,0.01613066593805949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,1536,0.021368889345063105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,768,0.01672266589270698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,65536,1.1148817274305556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,1024,0.013667555318938361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,768,0.012551999754375882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,768,0.013411555853154926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,512,0.011342222491900126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,256,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,512,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,3584,0.06263111035029094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,512,0.011844444606039258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,128,0.006017777654859755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,64,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,256,0.010225777824719747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,51200,128,0.010306666294733683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,51200,32,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,256,0.010669333239396414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,3584,0.04009600149260627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,128,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,64,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,32,0.01313244468635983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,65536,0.3265804449717204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,51200,0.25112888548109263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,16384,0.08602311213811238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,65536,0.22862577438354492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,51200,0.17524977525075278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,65536,0.14205066363016763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,51200,0.11201866467793782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,51200,1024,0.01703288820054796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,16384,0.06160089042451647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,10240,0.0548017786608802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,16384,0.042035556501812406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,12288,0.03285244438383315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,8192,0.04524888926082187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,7168,0.039692444933785334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,12288,0.046498666206995644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,10240,0.02805066770977444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,8192,0.02384177843729655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,10240,0.0399582220448388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,6144,0.03469244307941861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,5120,0.030238221089045208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,8192,0.03456799851523505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,7168,0.02125599980354309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,4096,0.02566844390498267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,7168,0.030003554291195337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,6144,0.019262222780121695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,3584,0.022230222821235657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,5120,0.01687111126052009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,6144,0.02663466665479872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,3072,0.0195768889453676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,4096,0.016524444023768108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,5120,0.022919111781650122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,2560,0.01646844380431705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,4096,0.019479110836982727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,3584,0.01755466726091173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,3072,0.012800888882742988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,2560,0.012067555553383298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,3072,0.01565511193540361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,2560,0.013840888937314352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,1536,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,1024,0.008355555435021719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,2048,0.011047999891969891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,12288,0.06411377588907878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,2048,0.012478222449620565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,1536,0.00965244405799442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,1024,0.00831911133395301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,1536,0.010768888725174798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,512,0.0052826665341854095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,1024,0.009296889106432596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,768,0.007669332954618666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,256,0.0034924443397257063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,768,0.008035555481910706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,128,0.00314666661951277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,512,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,256,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,512,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,64,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,32,0.0031582222630580268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,128,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,256,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,128,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,64,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,16384,32,0.007715555528799693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,65536,0.24727911419338652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,16384,3584,0.014170666535695394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,2048,0.013256000147925483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,16384,0.06378933456208971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,51200,0.186919993824429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,12288,0.04950399862395393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,65536,0.125965330335829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,51200,0.1349102258682251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,16384,0.037880887587865196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,10240,0.04136711027887132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,51200,0.0994302233060201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,12288,0.02951288885540432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,8192,0.03484355409940084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,12288,0.036884443627463445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,10240,0.025397333833906386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,10240,0.031911111540264554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,7168,0.02994399931695726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,8192,0.02164622147878011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,8192,0.027445332871543035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,6144,0.026360000173250835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,7168,0.019631110959582858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,5120,0.0229093333085378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,7168,0.024255111813545227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,6144,0.017975111802419026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,4096,0.02015911042690277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,6144,0.021575111481878493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,5120,0.016043555405404832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,3584,0.016840888394249808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,5120,0.01828533411026001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,65536,0.17655111683739555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,4096,0.014538667268223233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,3072,0.014479110638300577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,4096,0.01594311164485084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,2560,0.01237511138121287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,3584,0.013415111435784234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,16384,768,0.0069013333155049225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,3584,0.014469333820872836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,3072,0.012952889005343119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,2048,0.0107351111041175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,3072,0.013236444857385425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,1536,0.008730666504965888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,2560,0.011457777685589261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,1024,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,2048,0.010103999740547603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,1536,0.008919110728634728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,2048,0.011044444309340583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,768,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,1536,0.009521777431170145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,1024,0.007790221936172909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,1024,0.00814577771557702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,768,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,256,0.0032248888164758682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,768,0.007003555695215861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,512,0.007206222249401941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,128,0.0028906667398081887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,512,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,256,0.006637333167923822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,64,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,256,0.006260444306664997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,32,0.0028479999552170434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,12288,128,0.006654222392373615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,128,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,64,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,65536,0.20511644416385225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,32,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,16384,0.048551999860339694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,51200,0.1562888887193468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,65536,0.1668942239549425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,16384,0.05314755439758301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,12288,2560,0.012067555553383298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,65536,0.12182845009697808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,12288,0.041498667663998075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,51200,0.12954933113522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,16384,0.036377777655919395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,51200,0.09577955802281697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,12288,512,0.004011555678314633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,8192,0.029504888587527808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,16384,0.04610222246911791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,10240,0.024420445164044697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,10240,0.030549334155188665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,8192,0.02091111077202691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,6144,0.02223644488387638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,8192,0.026206221845414903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,7168,0.019000889526473153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,7168,0.023014222582181294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,5120,0.01924711134698656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,6144,0.017539555827776592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,4096,0.016377778516875375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,6144,0.020217777954207528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,5120,0.015711999601787992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,3584,0.013939556148317126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,5120,0.01759733259677887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,4096,0.014188443620999655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,3072,0.012320888539155325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,4096,0.015140444040298462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,3584,0.013244444297419654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,3584,0.014140443669425117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,2560,0.010676444404655032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,3072,0.012159999873903064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,10240,0.034675555096732244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,3072,0.013052444491121503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,2048,0.009399111072222391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,2560,0.010868444210953183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,1536,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,12288,0.028315554062525432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,12288,0.035169776942994856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,1024,0.006254222244024277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,2560,0.011752000285519494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,2048,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,7168,0.02533866630660163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,768,0.004552888787455029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,512,0.003548444559176763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,1536,0.00851466672288047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,2048,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,1024,0.007681777907742395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,1536,0.008984000318580205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,256,0.0029333333174387612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,768,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,1024,0.007320000065697565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,128,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,768,0.006911110960774952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,512,0.006696888970004187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,64,0.002536000063021978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,256,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,512,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,10240,32,0.002817777709828483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,10240,128,0.006670222100284364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,256,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,128,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,65536,0.18203911516401503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,51200,0.14008977678087023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,64,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,10240,32,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,16384,0.047932446002960205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,65536,0.07159288724263509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,12288,0.037478221787346735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,16384,0.022690667046440974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,51200,0.0576853354771932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,10240,0.03110044532352024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,16384,0.03628888726234436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,8192,0.026351999905374315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,12288,0.027622222900390625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,10240,0.01646577815214793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,10240,0.023923556009928387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,7168,0.023010666171709698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,8192,0.014464888307783337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,8192,0.0206977774699529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,7168,0.013408888545301227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,7168,0.01776355504989624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,5120,0.017128000656763714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,6144,0.012509333590666452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,6144,0.016005333926942613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,4096,0.01424711114830441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,5120,0.011512888802422417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,5120,0.014308444327778287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,3584,0.011696000066068439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,4096,0.01090400003724628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,65536,0.12432977888319229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,3072,0.010679110884666443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,4096,0.012826667063766055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,3584,0.009533333281675974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,51200,0.0963217748536004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,2560,0.009284444153308868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,2048,0.007932444413503012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,3584,0.01185155577129788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,3072,0.008866666919655269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,1536,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,12288,0.018571555614471436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,3072,0.011008000208271874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,1024,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,2048,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,2560,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,2048,0.009000000026490953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,1536,0.006979555719428592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,1024,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,1536,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,6144,0.01997511088848114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,512,0.003536000019974179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,768,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,256,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,1024,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,128,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,768,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,512,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,64,0.0025128889828920364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,256,0.005984000033802456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,32,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,512,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,128,0.00600000015563435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,128,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,65536,0.1607955561743842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,32,0.005997333261701796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,51200,0.12709244092305502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,65536,0.12173867225646973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,65536,0.0688755578464932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,16384,0.04855288730727302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,8192,2560,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,51200,0.0944639974170261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,12288,0.03520088725619846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,8192,768,0.004206222378545337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,16384,0.022095110681321885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,51200,0.05615644322501289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,16384,0.03451022174623277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,10240,0.028823110792371962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,12288,0.017910222212473553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,8192,0.025060445070266724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,7168,0.021354667014545862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,12288,0.02660711109638214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,10240,0.015847111741701763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,8192,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,6144,0.018769777483410306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,10240,0.022950222094853718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,7168,0.01308444473478529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,8192,0.01959733333852556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,7168,0.017253332667880587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,6144,0.01591377788119846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,4096,0.013780444032616086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,64,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,5120,0.011158222125636207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,5120,0.013975999421543546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,3584,0.011075555450386472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,4096,0.009859555297427708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,3072,0.010022222167915767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,3584,0.009179555707507664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,4096,0.012587555580668978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,2560,0.008759111166000366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,3584,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,3072,0.008269333177142674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,2048,0.007620444728268518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,3072,0.010855110983053843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,2560,0.007732444339328342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,2560,0.009902222288979424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,1536,0.00638133328821924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,2048,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,1024,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,2048,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,1536,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,768,0.0041831110914548235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,1536,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,1024,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,5120,0.015965332587560017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,512,0.0035084444615576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,8192,256,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,1024,0.00664444433318244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,768,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,512,0.006225777582989798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,768,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,128,0.0028035555862718155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,512,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,256,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,64,0.0026577777332729767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,32,0.0028026666906144884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,128,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,256,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,128,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,64,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,65536,0.14236177338494196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,7168,32,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,51200,0.11358133951822917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,16384,0.042706665065553456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,51200,0.05379377802213033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,51200,0.09028177791171604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,16384,0.021320000290870667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,12288,0.03324177861213684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,16384,0.0329804453584883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,10240,0.0236471113231447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,12288,0.017416889468828838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,7168,6144,0.012139555480745105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,12288,0.025225778420766194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,10240,0.015453333655993143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,7168,0.017162667380438913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,10240,0.021948445174429152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,8192,0.013796444568369122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,6144,0.014906666345066495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,8192,0.019097778532240126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,7168,0.012800888882742988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,5120,0.013061333033773633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,7168,0.017083555459976196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,6144,0.011791110866599612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,4096,0.011277333729796939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,6144,0.01555288831392924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,5120,0.010716444088353051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,3584,0.009966221948464712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,5120,0.013786666923099093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,65536,0.11706933710310195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,4096,0.00902133352226681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,4096,0.012429333395428128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,3584,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,2560,0.007822222179836696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,3584,0.011347555451922946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,3072,0.008021333151393467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,3072,0.010694222317801582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,2048,0.007151111132568783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,2560,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,1536,0.0058559998869895935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,7168,256,0.003152888889114062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,2560,0.009382222261693742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,2048,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,1024,0.004176889028814104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,2048,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,1536,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,768,0.0038773334688610504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,1024,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,1536,0.007243555453088548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,512,0.0031644445326593188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,1024,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,768,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,256,0.0028248888750871024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,768,0.006077333456940121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,128,0.002465777720014254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,512,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,256,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,64,0.0022622222701708474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,256,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,65536,0.06672710842556424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,128,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,32,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,128,0.005406222409672207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,65536,0.11548266145918105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,64,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,6144,32,0.006001777946949005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,51200,0.0927377806769477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,8192,0.019904888338512845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,65536,0.11067289113998413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,65536,0.06410488817426893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,16384,0.03716444306903415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,12288,0.0292871097723643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,51200,0.08603377474678887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,51200,0.05169066786766052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,16384,0.020459555917316012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,10240,0.025960889127519395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,16384,0.0316826668050554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,6144,512,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,8192,0.017995556195576985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,10240,0.01516888870133294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,12288,0.024147555232048035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,10240,0.021163556310865615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,8192,0.013321778012646569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,6144,3072,0.008912000391218398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,7168,0.012406222522258759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,8192,0.01825066738658481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,7168,0.016540444559521146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,5120,0.012057777908113269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,6144,0.011439999772442712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,4096,0.01035911093155543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,6144,0.015190222197108798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,5120,0.010096888575288985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,3584,0.008990222381220924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,5120,0.013397333522637686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,4096,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,3072,0.00833511104186376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,4096,0.012154666913880242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,3584,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,2560,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,3584,0.011120888921949597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,3072,0.007845333053006066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,2048,0.006510222123728857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,3072,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,2560,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,12288,0.016805332567956712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,1536,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,2560,0.008615111311276754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,2048,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,1024,0.0042133331298828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,2048,0.00794400026400884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,768,0.003802666647566689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,1024,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,6144,0.013590222431553734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,512,0.003152888889114062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,1024,0.0063368889192740125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,768,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,256,0.0028444443725877335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,512,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,768,0.00608711110221015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,128,0.002464888824356927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,512,0.00572622236278322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,64,0.0021911110315057966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,256,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,32,0.002493333278430833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,128,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,64,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,65536,0.09183022048738267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,32,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,51200,0.07464444637298584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,65536,0.061928888161977134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,5120,7168,0.015519110692871941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,16384,0.030293332205878362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,1536,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,51200,0.05423555440372891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,51200,0.04961688982115852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,12288,0.02402666707833608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,16384,0.019671999745898776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,16384,0.021003555920388963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,10240,0.020752888586786058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,12288,0.016399110356966656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,8192,0.015146666102939181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,256,0.005636444522274866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,5120,128,0.005879111174080107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,10240,0.01461155547036065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,7168,0.013325333595275879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,10240,0.014816888504558139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,6144,0.011720000041855706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,8192,0.013470222552617392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,7168,0.011894221935007306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,5120,0.010438222024175854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,7168,0.012040889097584618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,6144,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,6144,0.011258666714032492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,4096,0.009310222334331935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,5120,0.009382222261693742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,3584,0.008103110724025303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,5120,0.010054222411579555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,65536,0.06991377804014418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,4096,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,3072,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,4096,0.009005332986513773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,3584,0.00832622249921163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,2560,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,5120,1536,0.007185777856243982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,3584,0.00835111074977451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,3072,0.008015111088752747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,2048,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,3072,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,1536,0.004576888763242298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,2560,0.007601777712504069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,2560,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,12288,0.016911110944218107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,1024,0.0034906665484110513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,1536,0.006826666494210561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,2048,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,768,0.0032151111712058387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,1536,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,1024,0.006320888797442119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,8192,0.013031110995345645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,512,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,256,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,768,0.006165333506133821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,1024,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,128,0.002492444382773505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,768,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,512,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,64,0.00221422231859631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,256,0.005698666804366642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,512,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,4096,32,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,128,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,65536,0.08742133114073013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,128,0.005012444323963589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,64,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,32,0.005526222288608551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,51200,0.0704444448153178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,65536,0.06704444355434842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,65536,0.060821334520975746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,16384,0.0277528895272149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,12288,0.02159022291501363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,51200,0.05264711048867968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,16384,0.01953866746690538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,10240,0.019889778561062284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,16384,0.020113777783181932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,4096,2048,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,12288,0.016156444946924847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,8192,0.0144177774588267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,12288,0.016176000237464905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,10240,0.014434667097197639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,10240,0.01422755585776435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,8192,0.012493333054913415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,6144,0.011228444675604502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,8192,0.013103110922707452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,7168,0.011362666885058085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,5120,0.00998488896422916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,7168,0.011746666497654386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,6144,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,4096,256,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,4096,0.008943110704421997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,6144,0.01092088884777493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,5120,0.009165333376990424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,3584,0.00776444458299213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,4096,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,5120,0.010000000397364298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,3072,0.007231999602582719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,4096,0.008543111383914948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,2560,0.0064275554484791225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,3584,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,2048,0.005503111001518037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,51200,0.048635555638207324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,2560,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,1536,0.004479111068778568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,2560,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,2048,0.007256000406212277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,1024,0.00349866681628757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,2048,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,1536,0.006654222392373615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,7168,0.0127697777416971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,768,0.0031582222630580268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,1536,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,512,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,1024,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,768,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,256,0.002562666725781229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,512,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,768,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,128,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,256,0.0058764442801475525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,64,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,256,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,128,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3584,32,0.0022088889446523455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,128,0.005048888838953442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,65536,0.07687911060121325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,64,0.00519466648499171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,3584,0.007693332930405934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,32,0.005384888913896348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,51200,0.06225599845250448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,3072,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,65536,0.06437155273225573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,16384,0.023416888382699754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,65536,0.05960355864630806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,12288,0.019516444868511625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,51200,0.0474639998541938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,51200,0.050345778465270996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,16384,0.019325332509146798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,10240,0.016872889465755887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,12288,0.01590488851070404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,8192,0.014583999911944071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,12288,0.01575911045074463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,10240,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,7168,0.012867555850081973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,10240,0.013882666826248169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,8192,0.011800000237094032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3584,512,0.005305777821275923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,6144,0.011161777708265515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,8192,0.0125946667459276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,7168,0.011320888996124268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,5120,0.00963644435008367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,7168,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,4096,0.008552889029184977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,6144,0.010671111444632212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,5120,0.009138666921191746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,3584,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,5120,0.009532444179058075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3584,3072,0.00792266676823298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,4096,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,3072,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,4096,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,2560,0.0063786668082078295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,3584,0.008231999973456064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,3584,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,2048,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,2560,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,3072,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,1536,0.004854222138722737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,2560,0.00674044465025266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,16384,0.01924711134698656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,2048,0.007239999870459239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,1024,0.003896000070704354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,2048,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,1536,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,768,0.0035004446076022256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,1536,0.006034666465388403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,512,0.003145777723855443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,1024,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,768,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,1024,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,256,0.002640888922744327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,768,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,512,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,128,0.0025128889828920364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,256,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,6144,0.010040889183680216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,512,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,64,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,3072,32,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,256,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,128,0.005642666584915585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,65536,0.06587466928693983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,128,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,64,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,3072,32,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,65536,0.057951112588246666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,16384,0.020917332834667627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,3072,3072,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,65536,0.061971558464898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,12288,0.017530666457282174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,51200,0.04647555616166857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,10240,0.015115555789735583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,16384,0.01889866590499878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,8192,0.013350221845838757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,16384,0.018888889087571036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,12288,0.015678221980730694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,7168,0.011704000333944956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,10240,0.013597333596812354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,12288,0.015460444821251763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,8192,0.011528000235557556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,6144,0.01054666688044866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,8192,0.012121777567598553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,7168,0.010440000229411654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,5120,0.009338666995366415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,6144,0.009822222093741098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,7168,0.011306666665607028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,4096,0.008220444122950235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,6144,0.01015822258260515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,5120,0.009010666774378883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,3584,0.00739555557568868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,5120,0.008966222405433655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,4096,0.008398222426573435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,3072,0.006773333168692059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,3584,0.008023111356629265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,4096,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,2560,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,3584,0.007625777688291337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,51200,0.05404355459743076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,2048,0.005029333134492238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,3072,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,2560,0.007347555624114142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,1536,0.0042239998777707415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,2560,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,2048,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,51200,0.04860088891453213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,2048,0.006568888823191325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,1024,0.003555555724435382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,768,0.003212444484233856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,1536,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,1536,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,1024,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,512,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,10240,0.013603555659453074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,768,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,1024,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,256,0.00250133333934678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,768,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,512,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,128,0.002161777681774563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,64,0.0021946666141351065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,256,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,512,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2560,32,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,128,0.005707555760939916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,256,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,65536,0.054024888409508594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,128,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,51200,0.043939554029040866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,64,0.004970666848950916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2560,32,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,16384,0.017235555582576327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,65536,0.05671111080381605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,65536,0.05901510847939385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2560,3072,0.007746666669845581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,12288,0.013976888524161445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,51200,0.04527022110091316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,10240,0.012050666742854647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,51200,0.04655199911859301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,8192,0.012039110892348819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,16384,0.018548443913459778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,12288,0.015048889650238885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,10240,0.012440000143316058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,7168,0.01054488867521286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,10240,0.013278221918476952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,8192,0.011046222514576383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,6144,0.009748444788985783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,8192,0.01183644433816274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,7168,0.01036266651418474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,5120,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,7168,0.010778666370444827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,6144,0.009649777577983009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,4096,0.007720888488822513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,6144,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,5120,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,3584,0.0070222218831380205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,5120,0.008646222452322641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,4096,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,3072,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,4096,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,3584,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,2560,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,3584,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,2048,0.004578666554556953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,2560,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,3072,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,1536,0.003900444342030419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,2560,0.006719999843173557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,16384,0.018216000662909616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,2048,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,1024,0.003248888792263137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,2048,0.006507555643717448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,1536,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,1536,0.005967999911970562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,1024,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,512,0.0025395556456512874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,768,0.005991999887757831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,1024,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,256,0.002431111203299628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,768,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,128,0.0021199999998013177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,512,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,256,0.005753777921199799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,512,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,64,0.0018604445374674266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,32,0.0019582222319311565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,256,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,128,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,65536,0.04153866569201151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,64,0.005058666484223472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,2048,32,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,3072,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,65536,0.0562302205297682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,16384,0.013776889277829064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,51200,0.044497778018315635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,51200,0.044907556639777295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,12288,0.012031111452314587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,2048,768,0.0028719999310043124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,16384,0.017606221967273288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,10240,0.01111822244193819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,16384,0.018188445104493033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,12288,0.013433777623706393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,12288,0.014780445231331719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,8192,0.009735999835862054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,10240,0.012400888734393649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,10240,0.013048888908492194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,7168,0.009320889082219865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,128,0.005624889085690181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,8192,0.010999110837777456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,6144,0.00884622252649731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,8192,0.01050222251150343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,7168,0.010273777776294285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,5120,0.00794488853878445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,7168,0.01036888857682546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,6144,0.009661333428488838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,6144,0.009290667043791877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,5120,0.008974221845467886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,5120,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,3584,0.0063591111037466265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,4096,0.00831288927131229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,4096,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,2048,12288,0.015093333191341825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,3072,0.00563822231358952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,3584,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,2560,0.005147555636035072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,3072,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,3584,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,2048,0.004549333204825719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,3072,0.007071110937330458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,2560,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,1536,0.0042364444169733255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,2560,0.006802666518423293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,2048,0.007120889094140794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,1024,0.003499555504984326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,2048,0.006468444648716185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,1536,0.006744888921578725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,768,0.003084444544381566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,1536,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,1024,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,51200,0.03330844309594896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,512,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,65536,0.055416001213921436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,768,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,4096,0.007644444704055786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,256,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,128,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,768,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,512,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,64,0.0021537778278191886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,256,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1536,32,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,512,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1536,128,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,65536,0.030528889762030706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,128,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,51200,0.024999999337726172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,64,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,32,0.005167110926575131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,65536,0.0549831125471327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,16384,0.011350221931934357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,65536,0.05391110976537069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,51200,0.04424000117513868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,51200,0.04309422108862135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,16384,0.016215999921162922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,10240,0.011412444214026133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,16384,0.017290666699409485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,8192,0.01019377758105596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,12288,0.013433777623706393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,12288,0.013568888935777875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,10240,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,7168,0.00942133367061615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,8192,0.01073333372672399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,10240,0.011993777420785693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,1024,0.00563822231358952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,6144,0.008741333252853816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,7168,0.01036533299419615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,5120,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,8192,0.010545777777830759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,4096,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,6144,0.009698666632175446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,5120,0.008989333278603023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,6144,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,3584,0.006503110958470239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1536,256,0.004977777600288391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,4096,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,5120,0.008627555436558193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,3072,0.0058613332609335584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,2560,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,4096,0.007744889292452071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,3584,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,2048,0.004573333180612988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,3072,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,3584,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,2560,0.007178666690985362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,1536,0.0038782221575578055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,3072,0.007128888534175024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,1024,0.0031840000301599503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,2560,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,2048,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,12288,0.012447111308574677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,768,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,1536,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,2048,0.006542222367392645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,512,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,1024,0.006177777631415262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,768,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,1024,0.005640888793600931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,256,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,512,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,768,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,128,0.0021093332519133887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,512,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,256,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,64,0.0018044445249769424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,1024,128,0.005533333453867171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,256,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,128,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,65536,0.025511110822359722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,7168,0.009952888720565373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,51200,0.020584889584117465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,64,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,32,0.005190222213665645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,65536,0.05439466569158766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,16384,0.01236977759334776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,51200,0.04367555512322319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,65536,0.052483555343416005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,12288,0.010924444430404238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,51200,0.04225155711174011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,16384,0.01607111096382141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,10240,0.00980000032318963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,16384,0.016864000095261466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,12288,0.013286222186353473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,8192,0.008746667040718926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,12288,0.01348533315791024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,10240,0.012124444047609964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,7168,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,1024,1536,0.00600622221827507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,8192,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,6144,0.00758844448460473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,7168,0.01014666673209932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,5120,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,6144,0.009380444056457942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,6144,0.009325332939624786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,1024,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,5120,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,5120,0.00851022203763326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,3584,0.0063973334100511335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,4096,0.008100444244013892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,4096,0.007901333272457123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,3584,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,3072,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,3584,0.00738666703303655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,2560,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,3072,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,3072,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,2048,0.004337777694066365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,2560,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,2560,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,10240,0.011990221838156382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,1536,0.0038284444146686127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,2048,0.006986666884687211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,1024,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,1536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,7168,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,768,0.00286666676402092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,1536,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,1024,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,4096,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,1024,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,256,0.0021955555097924336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,768,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,512,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,128,0.002159111201763153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,512,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,256,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,64,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,256,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,32,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,65536,0.018812444474962022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,64,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,32,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,51200,0.015140444040298462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,65536,0.051739556921852954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,65536,0.053745776414871216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,16384,0.010037333601050908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,8192,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,2048,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,51200,0.04070933328734504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,12288,0.00871377769443724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,51200,0.04181333382924398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,768,512,0.0024871110088295406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,10240,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,16384,0.016146666473812528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,12288,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,768,768,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,8192,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,7168,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,12288,0.013432000246312885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,10240,0.012103111379676394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,8192,0.011031111081441244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,10240,0.011714666253990598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,6144,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,7168,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,8192,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,5120,0.00794222205877304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,7168,0.009694221946928237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,4096,0.006798222247097228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,6144,0.00962755580743154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,768,128,0.004965333475006951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,5120,0.008956444760163626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,3584,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,5120,0.008341333104504479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,4096,0.008292444050312042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,3072,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,4096,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,2560,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,3584,0.007277333074145847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,3072,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,2048,0.004566222015354368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,3072,0.006984000404675801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,2560,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,1536,0.0038533334930737815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,2560,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,2048,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,16384,0.015867556134859722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,1024,0.003137777869900068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,1536,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,768,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,1536,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,512,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,1024,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,768,0.006228444476922353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,256,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,768,0.0052879999081293745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,512,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,128,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,6144,0.009138666921191746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,256,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,64,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,128,0.005542222410440445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,3584,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,512,32,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,128,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,65536,0.015230221880806817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,64,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,65536,0.046979556481043495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,32,0.004956444518433677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,65536,0.0506951113541921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,512,2048,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,16384,0.008418666819731394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,51200,0.03768622212939792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,12288,0.01311466677321328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,51200,0.03944799966282315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,16384,0.016068443655967712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,10240,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,16384,0.01610577768749661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,12288,0.013409777647919126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,8192,0.008608889248636033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,512,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,10240,0.012082666986518435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,12288,0.013379555609491138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,7168,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,6144,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,10240,0.011715555356608497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,512,256,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,5120,0.006810666786299811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,7168,0.010245333115259806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,8192,0.010405333505736457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,6144,0.009597332941161262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,4096,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,7168,0.009673777553770278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,6144,0.009022222624884712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,3584,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,5120,0.008987555901209513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,3072,0.0053119998839166425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,4096,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,3584,0.007809778054555257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,4096,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,2560,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,51200,0.013168888787428537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,3072,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,3584,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,3072,0.007071110937330458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,2560,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,1536,0.003571555432346132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,2048,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,2560,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,1024,0.003142222141226133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,1536,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,2048,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,768,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,1536,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,1024,0.006260444306664997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,512,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,768,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,1024,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,256,0.0021626665774318907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,512,0.0058462222417195635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,768,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,128,0.001981333312061098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,8192,0.010679999987284342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,512,0.005039111193683413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,256,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,128,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,32,0.0018071112119489247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,65536,0.013914666242069669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,128,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,5120,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,64,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,51200,0.012280888855457306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,256,32,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,65536,0.04658311274316576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,16384,0.010312000082598792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,256,2048,0.00423200014564726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,65536,0.04825777808825175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,12288,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,51200,0.037660443120532565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,10240,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,51200,0.03865955604447259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,16384,0.015849777393870883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,8192,0.006766222417354584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,12288,0.013415111435784234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,16384,0.016157334049542744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,12288,0.013068444199032254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,7168,0.006423111177153057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,8192,0.010786666638321347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,10240,0.0116462219092581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,6144,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,8192,0.010265777508417765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,5120,0.0064088888466358185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,7168,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,6144,0.009664889011118147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,4096,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,256,256,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,6144,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,3584,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,5120,0.00868888861603207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,5120,0.008282666405042013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,4096,0.008355555435021719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,3072,0.00527555536892679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,4096,0.007713778151406183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,3584,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,2560,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,3072,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,3584,0.007172444628344641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,2048,0.00416088890698221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,3072,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,2560,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,1536,0.003546666767862108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,2048,0.006666666517655055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,2560,0.006405333264006509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,1024,0.002966222249799304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,2048,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,1536,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,768,0.0028248888750871024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,1024,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,10240,0.012068444656001197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,512,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,1024,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,768,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,256,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,7168,0.010221333139472539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,128,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,256,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,512,0.005111111121045219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,64,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,128,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,128,32,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,128,0.004944889081848992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,65536,0.016872000363137987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,51200,0.014534221755133735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,64,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,16384,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,32,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,12288,0.005976000179847081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,65536,0.04827377862400479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,10240,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,51200,0.03844088978237576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,16384,0.015776889191733468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,12288,0.013013333082199097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,7168,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,1536,0.006141333530346553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,10240,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,6144,0.005884444548024072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,5120,0.006353777729802662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,8192,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,113,128,512,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,7168,0.009602666729026372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,6144,0.009030222064918941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,3584,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,256,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,5120,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,2560,0.00519822206762102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,4096,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,2048,0.004568888909286923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,3584,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,1536,0.0035075553589397003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,2560,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,2048,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,1024,0.0028151110228565005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,8192,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,1536,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,768,0.0025191110455327565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,512,0.0025048889219760895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,1024,0.0053173332578606075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,256,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,768,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,4096,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,128,0.0018595556418100994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,64,0.0017804443422291016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,256,0.004851555658711327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,32,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,64,3072,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,65536,0.016721778445773654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,128,0.004612444589535396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,64,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,51200,0.01441333360142178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,16384,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,32,0.004581333448489507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,12288,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,3072,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,10240,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,65536,0.047888888253106006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,8192,0.0053395554423332214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,51200,0.03878844446606106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,16384,0.015842666228612263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,12288,0.012749333348539142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,6144,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,10240,0.011632888681358762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,5120,0.006230222268237009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,8192,0.010045333041085137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,4096,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,7168,0.009390222529570261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,3584,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,6144,0.00889244427283605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,3072,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,5120,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,128,768,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,64,512,0.004969777746333016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,2560,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,4096,0.007378666765160031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,2048,0.004576888763242298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,1536,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,1024,0.0028151110228565005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,3584,0.007003555695215861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,768,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,3072,0.006696888970004187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,512,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,2560,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,2048,0.00639111093348927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,256,0.0021608887861172357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,1536,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,128,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,64,0.0018053334206342697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,32,0.001806222316291597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,768,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,512,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,256,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,128,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,64,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,113,32,7168,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,32,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,16384,0.3709662225511339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,16384,0.20295200082990858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,51200,0.6216666433546278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,12288,0.2776675489213732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,51200,1.1536426544189453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,113,32,1024,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,16384,0.13208444913228354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,10240,0.23689866065979004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,51200,0.39721422725253636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,10240,0.13034133116404215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,12288,0.10076089037789239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,8192,0.19271911515129936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,10240,0.08842311302820842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,16384,0.2965928978390164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,8192,0.11097600062688191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,7168,0.17064800527360705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,12288,0.1507377756966485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,8192,0.07030755281448364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,7168,0.09192888604270087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,6144,0.1447991132736206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,7168,0.06121955315272013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,8192,0.14595645003848606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,6144,0.07928266790178087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,6144,0.05337511168585884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,5120,0.0666746695836385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,7168,0.13489066229926214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,6144,0.11400443977779812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,4096,0.09917244646284316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,5120,0.04822311136457655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,12288,0.22018755806816948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,10240,0.18496533234914145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,4096,0.05480622251828512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,4096,0.04014222158326043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,3584,0.08744711346096462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,3072,0.07621777719921537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,4096,0.08095200194252862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,3584,0.03565866748491923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,3584,0.049433777729670204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,5120,0.12238755491044785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,3584,0.07213866710662842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,3072,0.030812445614073012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,2560,0.06387111213472155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,3072,0.042824887567096286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,2048,0.05156977640257942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,3072,0.06480355395211114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,2560,0.03744622071584066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,2560,0.02702666653527154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,2560,0.05749155415429009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,1536,0.040616889794667564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,2048,0.023405333360036213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,5120,0.09783644146389431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,2048,0.03237955437766181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,2048,0.049066666099760264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,1536,0.025695110360781353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,1536,0.01924711134698656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,768,0.02179644505182902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,1536,0.041927112473381885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,1024,0.015792000624868605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,1024,0.020635555187861126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,512,0.015133332875039844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,1024,0.035522666242387556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,768,0.01419911119672987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,768,0.017196445001496207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,256,0.009725333087974125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,768,0.029771556456883747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,512,0.012562666502263812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,512,0.013631111217869652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,128,0.007160888777838812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,512,0.025553777813911438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,256,0.011376889215575324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,256,0.0122417774465349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,256,0.020949333906173706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,64,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,65536,128,0.011360000405046674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,32,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,128,0.01131377783086565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,128,0.016215110818545025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,65536,1024,0.02869066596031189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,64,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,65536,32,0.0151591118839052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,65536,0.6315911081102159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,51200,0.4847715695699056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,65536,51200,0.9180942111545138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,65536,1.1053137249416776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,51200,0.866554684109158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,16384,0.2815573215484619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,16384,0.16097688674926758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,65536,0.41217777464124894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,16384,0.1054240001572503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,51200,0.3203173213534885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,12288,0.1195182270473904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,10240,0.16558310720655653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,10240,0.10142399867375691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,12288,0.21269422107272676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,10240,0.06789244545830621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,16384,0.23515378104315865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,8192,0.13286756144629583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,12288,0.1800168885125054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,8192,0.0862942205535041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,7168,0.1163288884692722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,12288,0.08136000235875447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,8192,0.05557688739564684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,10240,0.15427377488878039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,7168,0.0729422238137987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,6144,0.09989333152770996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,8192,0.12175110975901286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,7168,0.04973511232270134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,6144,0.06362044148974948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,6144,0.04349244303173489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,7168,0.1082693338394165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,5120,0.053928888506359525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,4096,0.07102844450208876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,5120,0.038535998927222356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,6144,0.09445777866575454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,3584,0.06125422318776449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,4096,0.03384711013899909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,5120,0.0818364421526591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,4096,0.06869155830807157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,3072,0.05379022161165873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,3584,0.02921688887808058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,3584,0.06231377522150675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,3072,0.03531466590033637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,51200,0.7416933377583822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,5120,0.08499111069573297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,2560,0.04576711191071404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,3072,0.025851555996470984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,3072,0.05618221892250908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,2048,0.037865777810414634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,2560,0.02349600030316247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,2560,0.032607999112870954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,1536,0.029458665185504492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,2048,0.01977155605951945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,2560,0.04933866527345446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,2048,0.027071111732059058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,1024,0.02216800053914388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,4096,0.04461866617202759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,2048,0.04332000017166138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,1536,0.017859554953045316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,1536,0.02161600026819441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,768,0.016399110356966656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,1536,0.03654844562212626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,1024,0.0137155552705129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,3584,0.04006844427850511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,1024,0.017608889275126986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,512,0.011345778074529437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,1024,0.03028622269630432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,768,0.012459555433856117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,256,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,768,0.014953777194023132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,512,0.011287111375066968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,768,0.026887999640570745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,512,0.01184977756606208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,128,0.0058346668051348785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,256,0.010059555371602377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,512,0.022272000710169475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,256,0.01847288840346866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,64,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,128,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,51200,128,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,51200,32,0.00453244439429707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,128,0.014479999740918478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,32,0.013088888592190213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,64,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,65536,0.33655378553602433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,65536,0.22763376765780977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,51200,65536,0.956216918097602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,51200,0.2612488799624973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,65536,0.1399093336529202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,16384,0.08631200260586208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,51200,256,0.010753778119881948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,51200,0.11286932892269558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,12288,0.06688888867696126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,16384,0.06159377760357327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,16384,0.04164977868398031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,16384,0.059902224275800914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,12288,0.04626489016744825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,10240,0.05528888768619961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,12288,0.04734844300482008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,8192,0.04722311099370321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,10240,0.027952889601389568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,10240,0.03974044322967529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,51200,0.16040444374084473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,65536,0.20877244737413195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,8192,0.02392533255947961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,10240,0.041725334193971425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,8192,0.034607112407684326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,51200,0.17585421933068168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,6144,0.03446933295991685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,8192,0.03714222378200955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,7168,0.021472889516088698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,7168,0.029804444975323145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,5120,0.03021777669588725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,6144,0.026787555880016748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,6144,0.02013777858681149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,7168,0.032791998651292585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,4096,0.025018667181332905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,6144,0.03011466728316413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,5120,0.0176862229903539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,5120,0.023169777459568445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,3584,0.02193244463867611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,5120,0.02639466689692603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,12288,0.03198844525549147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,4096,0.019508444600635104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,3072,0.019645333290100098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,4096,0.023503111468421087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,3584,0.017644445101420086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,2560,0.01668177710639106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,3584,0.02204622162712945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,3072,0.01311377767059538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,3072,0.015678221980730694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,2048,0.013408888545301227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,2560,0.0120666664507654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,7168,0.03945511248376634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,2560,0.013909333282046847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,1536,0.010669333239396414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,2560,0.01794577803876665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,2048,0.011026666396194033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,2048,0.012487110992272695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,1024,0.008542222281297049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,2048,0.01681155628628201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,1536,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,1536,0.015225778023401896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,768,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,1024,0.008041777544551426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,1024,0.009543110926946005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,4096,0.015076445208655464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,512,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,1024,0.014085332552591959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,768,0.00794222205877304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,3584,0.014279999666743808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,768,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,512,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,512,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,512,0.010687111152542962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,128,0.003165333428316646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,256,0.006559999866618051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,256,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,128,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,1536,0.010008888940016428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,64,0.0028906667398081887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,128,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,128,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,64,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,16384,32,0.007836444510353936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,768,0.007677333222495184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,256,0.0035137778355015647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,65536,0.24739556842380098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,16384,3072,0.02010311186313629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,65536,0.17551822132534453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,65536,0.12553155422210693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,51200,0.13508444362216526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,16384,32,0.0031431110368834603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,16384,0.06399289104673597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,51200,0.09983733627531265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,16384,0.03699822227160136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,16384,0.048539555735058255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,12288,0.04968622326850891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,16384,0.05296711126963297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,65536,0.17385689417521158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,10240,0.04114400015936957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,12288,0.03675466775894165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,51200,0.13435733318328857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,16384,256,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,51200,0.18681155310736763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,12288,0.029493331909179688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,8192,0.03513777918285794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,12288,0.04203022188610501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,7168,0.030138668086793687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,10240,0.025495999389224585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,10240,0.03179288903872172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,8192,0.022065778573354084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,6144,0.026379555463790894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,8192,0.027441778116756018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,10240,0.03672000103526645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,7168,0.01960355540116628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,5120,0.022742221752802532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,8192,0.03307911091380649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,6144,0.01831999917825063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,7168,0.029374221960703533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,6144,0.021576889687114294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,4096,0.01995822290579478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,5120,0.016192888220151264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,6144,0.027281777726279363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,5120,0.018562666243977018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,3584,0.016528000434239704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,4096,0.014459555347760519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,5120,0.023936000135209825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,3072,0.01465777721669939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,4096,0.015840000576443143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,4096,0.02124799953566657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,3584,0.014514666464593677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,3584,0.013449778159459433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,3584,0.02014311154683431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,3072,0.012540444731712341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,2048,0.010572444233629437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,3072,0.018607111440764535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,2560,0.011730666789743634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,2560,0.012074666718641916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,1536,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,2560,0.017217778497272067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,2048,0.010222222242090438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,2048,0.01108266661564509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,1024,0.007096889118353526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,2048,0.015980444020695157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,1536,0.009010666774378883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,7168,0.024234667420387268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,768,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,1536,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,1536,0.014469333820872836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,1024,0.007821333077218797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,512,0.0042160000238153665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,768,0.007335111498832703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,1024,0.01276088919904497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,768,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,256,0.0038728887836138406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,512,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,512,0.007003555695215861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,2560,0.012291555603345236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,512,0.010330666270520952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,256,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,128,0.0031973332580592898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,3072,0.01330311099688212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,256,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,64,0.0028560000161329904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,256,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,12288,128,0.006650666809744305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,128,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,128,0.008360000120268928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,64,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,32,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,65536,0.20215645101335314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,51200,0.1564168930053711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,65536,0.1673262251747979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,12288,1024,0.008276444342401292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,65536,0.12008444468180339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,16384,0.05353688862588671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,51200,0.12915555636088052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,12288,768,0.011545778148704104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,51200,0.09610933065414429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,16384,0.04636977778540718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,12288,0.042259554068247475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,16384,0.03633600142267015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,12288,32,0.003165333428316646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,16384,0.05024266574117872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,10240,0.03458044595188565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,12288,0.03531733486387465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,12288,0.029215110672844782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,8192,0.029751999510659113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,65536,0.15994044144948324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,51200,0.1266097757551405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,12288,0.04072533382309808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,10240,0.024324445260895625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,7168,0.025384889708624944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,10240,0.030559112628300984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,8192,0.020921778347757127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,10240,0.035451554589801364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,6144,0.022279111875428095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,8192,0.0265600002474255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,8192,0.03127110997835795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,5120,0.019254222512245178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,7168,0.02828622195455763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,6144,0.017456000049908955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,6144,0.020230222079488967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,4096,0.016237333416938782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,5120,0.015805333852767944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,5120,0.017637333936161466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,5120,0.022671111755900916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,3584,0.01401422255569034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,4096,0.01420088940196567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,4096,0.015092444088723926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,4096,0.021021333005693223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,3072,0.012252444194422828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,3584,0.014150222142537435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,3584,0.013410666750537025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,3584,0.01953155630164676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,3072,0.012327999704413943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,2560,0.010337777435779572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,3072,0.012835555606418185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,7168,0.01935822268327077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,3072,0.017954667409261067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,2048,0.009121778110663096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,7168,0.022915555371178523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,2560,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,2560,0.010836444795131683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,1536,0.007583111524581909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,2048,0.009367110828558603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,2048,0.010836444795131683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,6144,0.026143110460705225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,1024,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,2048,0.01572800013754103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,1536,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,768,0.004783111313978831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,1536,0.0090844440791342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,1536,0.014527999692493014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,1024,0.0076657773719893555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,512,0.003726222034957674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,1024,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,768,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,1024,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,256,0.003116444374124209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,768,0.006649777707126405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,512,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,768,0.0110880004035102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,128,0.0028088889602157804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,512,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,256,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,64,0.0025351111673646504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,512,0.010772444307804108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,10240,32,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,10240,128,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,256,0.01032266683048672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,128,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,128,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,64,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,32,0.006896889044178857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,65536,0.17762577533721924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,10240,2560,0.01704355577627818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,51200,0.13981600602467856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,65536,0.12410844696892633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,16384,0.04798577891455757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,65536,0.07153777943717109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,51200,0.09577155775494045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,51200,0.05787822273042467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,12288,0.037258668078316584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,16384,0.023261333505312603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,16384,0.03508977757559882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,10240,0.031091554297341242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,12288,0.018598222070270114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,16384,0.04633511106173197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,12288,0.027479999595218237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,8192,0.025779555241266888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,12288,0.03752355443106757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,10240,0.016359999775886536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,10240,256,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,10240,0.023818666736284893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,7168,0.02295911146534814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,10240,0.03352800011634827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,8192,0.014529777897728814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,8192,0.02014488809638553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,6144,0.019965334071053397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,8192,0.02942755487230089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,7168,0.017902221944597032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,6144,0.012438221938080259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,7168,0.02702666653527154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,6144,0.016213332613309223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,6144,0.024312888582547505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,5120,0.014136888914638095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,5120,0.011546666423479715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,4096,0.014732443624072604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,5120,0.022138666775491502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,4096,0.010752889017264048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,51200,0.11379822095235188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,4096,0.012893333203262754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,65536,0.14190577136145696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,3584,0.013088888592190213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,4096,0.019711110326978896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,3072,0.011722666521867117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,3584,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,3584,0.01181688904762268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,3072,0.011077333655622272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,7168,0.013528889252079858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,2560,0.010302222437328762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,3584,0.018539556198649935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,3072,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,3072,0.017662222186724346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,2048,0.008642666869693333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,2560,0.008072000410821702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,5120,0.01703733371363746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,2048,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,2560,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,2048,0.009193778038024902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,2560,0.016175111134847004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,1024,0.005701333284378052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,2048,0.01553422212600708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,1536,0.007002666592597961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,768,0.004518222063779831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,1536,0.007480888730949826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,1024,0.00665155549844106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,1024,0.0069955554273393415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,1536,0.013585777746306526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,512,0.0037600000699361167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,1024,0.011395555403497485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,768,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,768,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,256,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,768,0.010681777364677854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,512,0.005987555616431766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,128,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,512,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,512,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,256,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,256,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,256,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,32,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,8192,128,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,8192,128,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,64,0.005657777604129579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,65536,0.1596897840499878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,1536,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,32,0.0063075555695427795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,51200,0.126364443037245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,65536,0.06889244582917955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,51200,0.09440710809495713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,51200,0.05574578046798706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,16384,0.03424711028734843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,51200,0.1131226619084676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,65536,0.14101955625745985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,8192,64,0.002540444541308615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,16384,0.022231111923853557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,16384,0.04548266530036926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,8192,128,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,12288,0.01794666714138455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,12288,0.026328000757429335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,12288,0.037286221981048584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,10240,0.022980444961124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,8192,0.025036444266637165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,10240,0.015855110353893705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,65536,0.12340354919433594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,7168,0.021459556288189355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,10240,0.032920890384250216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,8192,0.014075555735164218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,16384,0.04863466819127401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,6144,0.018577777677112155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,8192,0.019745777050654095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,8192,0.028639998700883653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,7168,0.017469333277808297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,5120,0.015885333220163982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,7168,0.025735111700163946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,6144,0.012083555261294046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,6144,0.01595288846227858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,4096,0.013910222384664746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,6144,0.023856000767813787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,5120,0.011133333047231039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,12288,0.03441333439615037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,5120,0.014130666851997375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,3584,0.011072888970375061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,10240,0.028853333658642236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,5120,0.021527111530303955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,4096,0.010002666877375709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,4096,0.012677333421177335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,3072,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,4096,0.01942222151491377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,3584,0.009320889082219865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,3584,0.011738667057620155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,3072,0.010757333702511258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,3584,0.018294221825069852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,3072,0.017190222938855488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,2048,0.0075795559419526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,2560,0.009768889182143742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,7168,0.013165333204799227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,2560,0.015824889143308003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,1536,0.006333333336644703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,2048,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,2048,0.015264888604482016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,1536,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,1536,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,1024,0.005197333378924264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,1536,0.012798222402731577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,1024,0.006773333168692059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,1024,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,768,0.004226666771703296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,1024,0.011188444164064197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,3072,0.008441777692900764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,768,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,2560,0.00777599960565567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,768,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,768,0.011044444309340583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,256,0.0032017777363459268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,512,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,512,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,2048,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,128,0.0028373334142896864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,512,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,256,0.0058622223635514574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,256,0.00926577796538671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,64,0.0027795556104845474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,32,0.0027946666296985415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,7168,128,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,64,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,2560,0.008631111019187504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,65536,0.13145689169565836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,7168,512,0.003600888782077365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,65536,0.1156888935301039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,51200,0.10610577795240615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,65536,0.06682578060362074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,51200,0.05385066734419929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,51200,0.08957688675986396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,256,0.005648889061477449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,16384,0.041519112057156034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,65536,0.13998577329847547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,128,0.00545155546731419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,12288,0.03422044383154975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,16384,0.03289866778585646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,51200,0.11240711477067734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,16384,0.021136000752449036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,10240,0.02342133389578925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,16384,0.04530310961935255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,12288,0.025408888856569927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,12288,0.017308443784713745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,8192,0.020208888583713107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,10240,0.01535999940501319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,12288,0.03693600164519416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,10240,0.021754667162895203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,10240,0.032391111056009926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,8192,0.013819555441538492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,8192,0.018965333700180054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,6144,0.014814222852389017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,8192,0.02751733362674713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,7168,0.012793777717484368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,7168,0.02555911077393426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,6144,0.011742221812407175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,6144,0.015536889433860779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,5120,0.013097777962684631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,6144,0.023199111223220825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,4096,0.011395555403497485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,5120,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,5120,0.013775111072593264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,5120,0.021012443635198805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,3584,0.009889778163697984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,4096,0.012416000167528788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,4096,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,7168,32,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,3072,0.008991999758614434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,4096,0.019129777948061626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,3584,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,7168,0.01715555621518029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,3584,0.011380444798204633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,2560,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,3584,0.017823111679818895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,3072,0.008384889198674096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,2048,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,3072,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,3072,0.017228444417317707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,2560,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,7168,0.01700177788734436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,2560,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,2560,0.01585155559910668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,2048,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,2048,0.008071111308203803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,1024,0.004226666771703296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,2048,0.014129777749379476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,1536,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,7168,128,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,768,0.0038577777643998465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,1536,0.012119111087587146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,1024,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,1024,0.011005333728260465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,512,0.0034844444857703317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,768,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,768,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,768,0.010741333166758219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,256,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,512,0.0058746664888328975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,512,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,512,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,128,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,256,0.005884444548024072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,256,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,1536,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,256,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,128,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,32,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,128,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,6144,128,0.007378666765160031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,1536,0.007290666302045186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,64,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,65536,0.11343555980258518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,6144,1024,0.00638755535085996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,51200,0.09014933639102513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,65536,0.06417866547902425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,51200,0.08627111381954616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,51200,0.051742222574022084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,16384,0.03710222244262695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,65536,0.13979821734958225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,51200,0.11223555935753717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,12288,0.02863022353914049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,16384,0.020337777005301583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,16384,0.03179555469089084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,6144,64,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,10240,0.02513422237502204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,16384,0.04444977641105652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,12288,0.0169813334941864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,12288,0.024285333024130926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,8192,0.018192888961897958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,10240,0.014950222439236112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,10240,0.021220443977249995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,12288,0.03642400105794271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,7168,0.015476443701320224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,10240,0.030905776553683813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,6144,32,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,8192,0.013378666506873237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,6144,0.013633777697881063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,8192,0.018236445056067575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,7168,0.012416889270146688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,8192,0.027109333210521277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,7168,0.016607999801635742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,5120,0.011881777809725868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,7168,0.024977778395016987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,6144,0.01144088887506061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,6144,0.01519733336236742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,5120,0.010011555420027839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,6144,0.023014222582181294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,3584,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,5120,0.02095288866096073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,4096,0.01204711116022534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,4096,0.019116444720162284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,3072,0.008468444148699442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,3584,0.011254222028785281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,3584,0.008310221963458592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,3584,0.01755199995305803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,2560,0.007415999968846639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,3072,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,3072,0.010494222243626913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,3072,0.01687733332316081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,2048,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,2560,0.007450666692521837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,2560,0.009209777745935652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,2560,0.014787554740905762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,65536,0.11027199692196316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,1536,0.005205333232879639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,2048,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,2048,0.007679999702506595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,1024,0.004337777694066365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,4096,0.010314666562610203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,2048,0.014302222265137566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,1536,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,768,0.003842666745185852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,1536,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,5120,0.013444444371594323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,4096,0.00867377801073922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,1024,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,1536,0.01273333364062839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,512,0.0031902222997612427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,1024,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,768,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,256,0.0028151110228565005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,1024,0.01107911103301578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,768,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,512,0.005976000179847081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,128,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,768,0.010671111444632212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,512,0.005818666683302985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,256,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,256,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,512,0.010334221853150262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,64,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,256,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,5120,128,0.005728888842794631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,128,0.005320888840489917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,5120,128,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,64,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,5120,32,0.0057813334796163775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,51200,0.0746248894267612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,65536,0.0697279969851176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,65536,0.06189333067999946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,16384,0.030470222234725952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,51200,0.0543502238061693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,51200,0.049674666590160794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,65536,0.13992799652947321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,12288,0.024316444993019104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,16384,0.021031111478805542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,51200,0.11176799403296576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,16384,0.01963200006220076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,10240,0.021183111601405676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,16384,0.04388711187574598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,12288,0.01623288955953386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,12288,0.016520889268981088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,8192,0.015523556205961438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,12288,0.0355751117070516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,10240,0.01455200049612257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,10240,0.014817777607176038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,7168,0.013432000246312885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,10240,0.030576888057920668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,65536,0.09235466851128472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,8192,0.012770666844315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,6144,0.01202044470442666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,8192,0.01371022231049008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,7168,0.011809777882364062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,8192,0.026983999543719824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,7168,0.011954666839705573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,5120,0.010320888625250923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,7168,0.024700444605615404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,6144,0.010689777632554373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,6144,0.011367999845080905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,4096,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,6144,0.022663111488024395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,5120,0.009729777773221334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,5120,0.010214221974213919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,3584,0.008204444415039485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,5120,0.020800888538360596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,4096,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,4096,0.008967111508051554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,4096,0.01851377718978458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,3584,0.008197333249780867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,3584,0.008350222474998897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,2560,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,3584,0.017304889029926725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,3072,0.007792889244026608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,2048,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,5120,32,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,3072,0.016192888220151264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,2560,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,1536,0.004855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,2560,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,2560,0.014183110660976835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,2048,0.007074666519959767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,1024,0.0036275556517971885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,2048,0.006568000134494569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,1536,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,1536,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,768,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,1536,0.0124195557501581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,1024,0.011031111081441244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,512,0.002806222273243798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,768,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,768,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,768,0.010767999622556897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,3072,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,256,0.0026364444444576898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,512,0.005877333382765452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,512,0.00536088893810908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,128,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,512,0.010130667024188572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,3072,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,256,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,64,0.0021626665774318907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,256,0.009713778065310584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,128,0.005092444519201915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,4096,32,0.0022133334229389825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,128,0.005706666658322017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,65536,0.08205511172612508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,128,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,64,0.0052977779673205475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,32,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,4096,2048,0.013481777575280933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,65536,0.0664186676343282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,65536,0.06085777944988675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,4096,1024,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,16384,0.028007109959920246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,51200,0.048623998959859215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,65536,0.13969243897332087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,16384,0.020048000746303134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,51200,0.11137510670555963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,16384,0.019482667247454327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,12288,0.02165777815712823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,4096,256,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,16384,0.043772445784674756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,12288,0.016188444362746347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,12288,0.016186666157510545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,12288,0.03493955400254991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,8192,0.015001777145597668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,10240,0.014162666267818876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,10240,0.03035911255412632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,51200,0.06657689147525363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,7168,0.012816888590653738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,8192,0.012700444294346703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,8192,0.013014222184816996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,8192,0.026879999372694228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,6144,0.011629333098729452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,7168,0.011417778001891242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,51200,0.05197244551446703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,7168,0.011724444727102915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,5120,0.00998222248421775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,7168,0.024460444847742718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,6144,0.010053333308961656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,6144,0.010775999890433418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,4096,0.008950221869680617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,6144,0.02295200030008952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,5120,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,10240,0.019729778170585632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,3584,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,5120,0.020282665888468426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,4096,0.008360889222886827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,4096,0.008568888737095727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,3072,0.007123555574152205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,4096,0.018838221828142803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,3584,0.008026666939258575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,3584,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,2560,0.006370666540331311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,3584,0.01716088917520311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,3072,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,3072,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,2048,0.005415110952324337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,3072,0.015749333633316886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,2560,0.0069066666894488865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,1536,0.0042408886882993905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,2560,0.014145778285132514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,2048,0.007138667007287343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,2048,0.00654666663871871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,1024,0.0035004446076022256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,2048,0.012805333567990197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,1536,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,1536,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,768,0.0031724443866146933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,1536,0.0118231111102634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,5120,0.00998222248421775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,1024,0.006467555546098285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,1024,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,512,0.002982222164670626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,1024,0.0110880004035102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,768,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,256,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,512,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,10240,0.014463999205165438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,768,0.010398222340477837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,128,0.002507555608948072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,256,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,256,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,256,0.009323555562231276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,128,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,128,0.004955555415815777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,2560,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,64,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3584,32,0.0022026666750510535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,65536,0.07182577583524916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,64,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,32,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,65536,0.06368177466922335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,51200,0.058780444992913135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,65536,0.05992533100975884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,65536,0.13927911387549505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,51200,0.05016888843642341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,16384,0.023295111126369897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,51200,0.04763466782040066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3584,512,0.0052915554907586836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,12288,0.019518221418062847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,512,0.01034400032626258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,51200,0.11175466908348931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,16384,0.019150222341219585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,16384,0.019613333874278598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,10240,0.01664977769056956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,12288,0.015784000356992085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,16384,0.043291555510626904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,12288,0.015666666958067153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,8192,0.014359111587206522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,12288,0.03448266784350077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,10240,0.014137778017255994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3584,128,0.0074133334888352295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,10240,0.013817778064144982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,7168,0.013372444444232516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,10240,0.030118223693635728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,8192,0.011939555406570435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,6144,0.012061333490742577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,7168,0.011018666956159802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,7168,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,7168,0.02461155586772495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,5120,0.009669333696365356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,6144,0.009758222434255812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,6144,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3584,768,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,4096,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,6144,0.022259554929203455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,5120,0.008982222113344405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,5120,0.009333333207501305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,3584,0.00777511133088006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,5120,0.020206222931543987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,4096,0.008301333420806462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,4096,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,3072,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,4096,0.018028444714016385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,3584,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,3584,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,2560,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,3584,0.01648977730009291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,3072,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,3072,0.015506666567590503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,2048,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,2560,0.007322666545708974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,8192,0.012436444560686747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,2560,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,1536,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,8192,0.026692445079485577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,2560,0.01407911048995124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,2048,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,1024,0.003891555385457145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,1536,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,2048,0.013056889176368713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,1536,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,768,0.0035253332720862497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,1536,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,1024,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,1024,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,512,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,1024,0.011011555790901184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,768,0.006220444622966979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,768,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,256,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,512,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,512,0.009659555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,3072,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,128,0.0024968888610601425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,256,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,256,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,256,0.009020444419648912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,64,0.0021271111650599376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,3072,32,0.002151111140847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,3072,128,0.0057857777509424425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,128,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,3072,128,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,65536,0.06018933322694567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,64,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,32,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,51200,0.04993866549597847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,65536,0.05846755372153389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,65536,0.06143555376264784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,16384,0.02056711084312863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,768,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,51200,0.048524445957607694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,51200,0.046216888560189136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,65536,0.14007288879818386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,12288,0.01676177812947167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,512,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,51200,0.11102755864461263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,16384,0.018967999352349173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,16384,0.018735110759735107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,10240,0.014485332700941296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,12288,0.015499555402331881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,12288,0.015351111690203348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,8192,0.012602667013804117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,10240,0.013074666261672974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,12288,0.034824887911478676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,10240,0.013764444324705334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,7168,0.011384888655609555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,10240,0.03057510985268487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,8192,0.011360888679822287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,6144,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,8192,0.012103999654452005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,3072,2048,0.006424888968467712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,7168,0.01130311108297772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,7168,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,5120,0.009455111291673448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,7168,0.024616888827747766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,6144,0.010081777969996134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,4096,0.008294222255547842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,6144,0.022456000248591106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,5120,0.00906844437122345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,5120,0.019535111056433786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,4096,0.008409777449236976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,4096,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,3584,0.007354666789372762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,4096,0.017509332961506315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,16384,0.0424924459722307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,3072,0.00686488880051507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,3584,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,3584,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,3584,0.016122667325867545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,2560,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,3072,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,3072,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,8192,0.026416889495319788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,3072,0.015270221564504834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,2048,0.005195555587609609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,2560,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,2560,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,1536,0.004221333397759331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,2048,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,2560,0.01426577733622657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,5120,0.009027555584907532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,2048,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,1024,0.0035413333939181436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,2048,0.013059555656380124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,1536,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,1536,0.006028444402747684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,768,0.0031893334041039148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,1024,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,1536,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,1024,0.00563733321097162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,512,0.0028560000161329904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,1024,0.011165333290894827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,768,0.005998222364319696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,768,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,256,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,768,0.010728889041476779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,128,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,512,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,512,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,256,0.005776888794369168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,512,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,64,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,256,0.004953777624501122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,128,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,65536,0.051510221428341336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,128,0.005061333378156026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,128,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,64,0.004955555415815777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2560,6144,0.009648000200589498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2560,32,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,51200,0.04231999980078804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,65536,0.056553781032562256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,65536,0.05815644396675957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,16384,0.01720177796151903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,51200,0.04518844352828132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,51200,0.04684622089068095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,65536,0.13938132921854654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,16384,0.018536888890796237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,16384,0.018265777164035372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,16384,0.042618665430280894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,12288,0.014941333068741692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,12288,0.015180443723996481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,12288,0.03432888785998026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2560,32,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,8192,0.011231111155615913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,10240,0.013097777962684631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,10240,0.01294311136007309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,10240,0.030421333180533514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,7168,0.010443555812040964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,8192,0.01107911103301578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,8192,0.011699555648697747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,8192,0.026339555780092876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,6144,0.010292444792058732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,7168,0.010446222292052375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,7168,0.010905777414639791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,7168,0.023953777220514085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,12288,0.014124444789356656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,6144,0.00965244405799442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,51200,0.11103466484281753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,6144,0.009575111170609793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,4096,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,6144,0.02195377813445197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,5120,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,10240,0.01202933324707879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,5120,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,3584,0.007231111327807109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,4096,0.008349333372380998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,4096,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,3072,0.006545777950021956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,4096,0.017482666505707633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,3584,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2560,256,0.009303111169073317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,2560,0.00544088871942626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,3584,0.016002666619088914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,3072,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,3072,0.007237333390447829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,2048,0.004535111288229625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,3072,0.014951111541854011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,2560,0.006656000183688269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,2560,0.014076444837782117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,5120,0.009005332986513773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,2048,0.006562666760550604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,2048,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,1536,0.0038844446341196695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,2048,0.012791111237472959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,1024,0.0031662223239739737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,1536,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,1536,0.006184000107977126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,1536,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,1024,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,768,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,5120,0.01977333260907067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,1024,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,512,0.002570666579736604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,1024,0.011013333168294696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,3584,0.007456888755162557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,768,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,768,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,256,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,512,0.0058666666348775225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,768,0.01071111112833023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,128,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,512,0.005323555734422471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,512,0.010012444522645738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,2560,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,64,0.001869333287080129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,256,0.005048888838953442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,2048,32,0.002009777766135004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,256,0.009663999908500248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,128,0.004966222163703707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,65536,0.04158488909403483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,64,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,2048,32,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,51200,0.033258666594823204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,65536,0.056523554854922824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,65536,0.05554844604598152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,65536,0.13948799504174128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,51200,0.04461600051985847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,16384,0.01346933344999949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,51200,0.044812444183561534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,51200,0.11061066389083862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,12288,0.012429333395428128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,16384,0.018062222335073683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,256,0.005662222289376789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,10240,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,12288,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,12288,0.014751111467679342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,2048,128,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,8192,0.009870222045315636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,12288,0.03386044502258301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,2048,128,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,10240,0.012328888807031842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,10240,0.012449777788586087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,7168,0.00943199959066179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,8192,0.011049778097205691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,10240,0.029887109994888306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,8192,0.011041777829329172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,6144,0.008841777841250101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,8192,0.02605244517326355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,7168,0.01033333357837465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,7168,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,5120,0.008251555263996124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,7168,0.023544889357354905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,6144,0.009670221971140968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,6144,0.009328888522254096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,4096,0.007876444194051955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,6144,0.021313778228229944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,5120,0.00902133352226681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,5120,0.008548444343937768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,3584,0.006555555595291986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,5120,0.019000889526473153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,16384,0.01720622181892395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,4096,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,3072,0.0057795556883017225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,4096,0.00794400026400884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,4096,0.01682222220632765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,3584,0.007272889216740926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,2560,0.00519822206762102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,3072,0.007699555820888943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,3584,0.015998222761683993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,3072,0.007069333559936947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,2048,0.004558222161398994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,3072,0.01477599971824222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,2560,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,1536,0.004181333300140169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,2560,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,2048,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,2560,0.013743110828929476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,2048,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,1024,0.003528888854715559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,2048,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,1536,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,1536,0.006040888941950268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,768,0.0031271111220121384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,1536,0.011810666984981961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,1024,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,512,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,1024,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,768,0.006063111126422882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,768,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,768,0.010517333116796283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,256,0.002455111179086897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,512,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,512,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,512,0.009685333404276106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,16384,0.04224710994296604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,128,0.0021066665649414062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,256,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,256,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,64,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,256,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,128,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,128,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1536,128,0.007603555917739868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,65536,0.030282666285832722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,64,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1536,32,0.005003555367390315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,51200,0.023735110958417256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,65536,0.054697778489854604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,65536,0.053957333167394005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,65536,0.13920177353752985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,51200,0.042816887299219765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,16384,0.011517333487669626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,51200,0.11018844445546468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,3584,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,16384,0.017716444200939603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,12288,0.012780444489585029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,16384,0.016198222835858662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,16384,0.04179111123085022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,10240,0.011432888607184092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1536,32,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,12288,0.013426666458447775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,8192,0.010319111247857412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,12288,0.014039999908871122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,10240,0.012153777811262341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,12288,0.03343377841843499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,10240,0.012048000262843238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,7168,0.009407999614874521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,10240,0.028767112228605483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,8192,0.010848888920413123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,8192,0.01036888857682546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,6144,0.009683555199040307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1536,1024,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,7168,0.010038221875826517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,5120,0.00832799987660514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,7168,0.009941333697901832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,7168,0.022849778334299724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,6144,0.009684444301658207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,6144,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,4096,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,6144,0.021291555629836187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,5120,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,5120,0.008294222255547842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,3584,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,5120,0.019260444574885897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,4096,0.00831288927131229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,4096,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,4096,0.01720266706413693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,3584,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,3584,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,2560,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,3584,0.015496888094478183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,3072,0.0070088886552386815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,3072,0.007667555577225155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,2048,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,3072,0.01483022173245748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,2560,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,2560,0.006646222124497096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,1536,0.004232888834344016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,2048,0.0069439998931354946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,2048,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,8192,0.02535733415020837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,1024,0.0033573332346147964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,2048,0.013099555340078143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,1536,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,51200,0.04416977696948581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,768,0.0031431110368834603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,1536,0.011805333197116852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,1024,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,1024,0.005624889085690181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,512,0.002625777696569761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,1024,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,768,0.006084444622198741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,3072,0.005542222410440445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,256,0.002475555572244856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,768,0.010151111417346531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,512,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,512,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,128,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,512,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,256,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,256,0.005192888693677055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,64,0.0018995555324686898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,256,0.009039999710188972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,1024,128,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,1024,32,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,128,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,2560,0.013784888717863293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,65536,0.025409777959187824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,1024,128,0.007682667010360294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,64,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,51200,0.020873778396182593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,32,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,1536,0.006119110931952794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,65536,0.05439555644989014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,65536,0.05281066563394335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,16384,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,65536,0.1383591095606486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,51200,0.04377511143684387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,12288,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,16384,0.016547555724779766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,16384,0.016008888681729633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,1024,768,0.005537777725193236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,16384,0.04199022385809156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,51200,0.11009866661495632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,10240,0.010047111246320937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,12288,0.013426666458447775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,12288,0.013339555925793119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,8192,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,10240,0.012089778151777057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,12288,0.03289599882231818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,10240,0.012079111403889127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,7168,0.00831822223133511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,10240,0.02886577778392368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,8192,0.010709332923094431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,8192,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,6144,0.007763555480374231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,8192,0.024717332588301763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,7168,0.01034044474363327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,7168,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,5120,0.006986666884687211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,7168,0.022648889157507155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,6144,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,6144,0.009307555854320526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,4096,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,6144,0.020578665865792167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,5120,0.008769777913888296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,3584,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,5120,0.008396444221337637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,5120,0.018953777021831937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,51200,0.042428443829218544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,4096,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,3072,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,4096,0.01683555543422699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,3584,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,3584,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,3584,0.016195555528004963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,3072,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,3072,0.01477777792347802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,2048,0.004544888933499654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,2560,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,2560,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,2560,0.013870221873124441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,1536,0.003882666842805015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,2048,0.006370666540331311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,2048,0.012727999852763282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,1536,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,1024,0.0031715554909573663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,1536,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,1536,0.0120328888297081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,768,0.0028568889117903183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,4096,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,1024,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,1024,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,512,0.002528000002106031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,1024,0.010672888822025724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,768,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,2560,0.004996444616052839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,768,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,256,0.0022479999396536085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,3072,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,512,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,128,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,512,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,256,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,256,0.005043555465009478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,512,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,64,0.0018880000958840053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,768,32,0.002128888956374592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,256,0.009572444690598382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,128,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,65536,0.019101333287027147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,768,2048,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,128,0.004945777770545748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,51200,0.015351111690203348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,128,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,64,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,768,32,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,65536,0.054077333874172635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,65536,0.05185777942339579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,51200,0.041335999965667725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,51200,0.04154933492342631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,12288,0.00867733359336853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,16384,0.015935111376974318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,51200,0.11018044418758816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,16384,0.04115022222201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,10240,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,12288,0.013535999589496188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,12288,0.013368888861603208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,12288,0.032800889677471586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,768,768,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,8192,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,10240,0.011737777955002256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,10240,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,10240,0.02812533246146308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,7168,0.008437333007653555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,8192,0.010852444503042432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,8192,0.01037688884470198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,8192,0.0251555558707979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,7168,0.010047111246320937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,6144,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,7168,0.009664889011118147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,7168,0.02236088944805993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,6144,0.009418666362762451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,65536,0.13868978288438585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,5120,0.007243555453088548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,6144,0.009092444347010719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,16384,0.016181333197487723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,4096,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,6144,0.021278222401936848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,5120,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,3584,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,5120,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,4096,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,4096,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,5120,0.018944889307022095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,3072,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,4096,0.016529776983790927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,3584,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,3584,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,2560,0.004942222187916438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,3584,0.015522667103343539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,3072,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,2048,0.004250666747490565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,3072,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,2560,0.007241778075695038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,2560,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,1536,0.0038453332251972626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,2560,0.013791110780504016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,2048,0.006912888752089606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,2048,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,16384,0.010206222534179688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,1024,0.0031511110977994073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,2048,0.013091555900043912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,1536,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,1536,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,768,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,1536,0.012119111087587146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,1024,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,512,0.0024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,1024,0.01071911139620675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,768,0.006207110981146495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,768,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,768,0.010339555641015371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,512,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,512,0.009703999592198266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,256,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,256,0.004975111120276981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,64,0.0018053334206342697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,256,0.0086986662613021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,32,0.0018577778504954444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,128,0.004981333182917701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,3072,0.014485332700941296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,65536,0.015269334117571512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,512,128,0.007256000406212277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,64,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,51200,0.01311911145846049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,512,32,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,65536,0.04698044392797682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,65536,0.050944887929492526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,16384,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,65536,0.13793333371480307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,1024,0.0063031112982167145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,51200,0.037688887781567044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,51200,0.03979644510481093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,12288,0.007290666302045186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,16384,0.015875554747051664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,16384,0.01605244477589925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,10240,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,16384,0.04084088736110263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,12288,0.01312888910373052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,512,128,0.0021582223061058256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,12288,0.03285600079430474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,8192,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,10240,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,10240,0.011744000017642975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,7168,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,512,128,0.005659555395444234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,10240,0.028742220666673448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,8192,0.01035377797153261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,6144,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,8192,0.024728000164031982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,7168,0.01033155537313885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,7168,0.009707555174827576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,5120,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,7168,0.023010666171709698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,6144,0.009663110805882348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,6144,0.009003555609120263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,6144,0.02047911120785607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,4096,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,5120,0.00868088917599784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,5120,0.008413333031866286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,51200,0.10921333233515422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,5120,0.01886133352915446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,3584,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,4096,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,4096,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,12288,0.013412444127930535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,3072,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,4096,0.01681511104106903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,3584,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,3584,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,2560,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,3584,0.015480000111791821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,3072,0.007351111206743453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,2048,0.00423200014564726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,8192,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,3072,0.014656888114081489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,2560,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,2560,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,2560,0.013819555441538492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,2048,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,2048,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,1024,0.003148444410827425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,2048,0.01276088919904497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,1536,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,1536,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,768,0.0027991111079851785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,1536,0.012082666986518435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,1024,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,1024,0.010434666441546546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,768,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,512,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,768,0.010688888529936472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,512,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,512,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,256,0.0021262222694026097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,512,0.009483555952707926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,3072,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,128,0.0018622221218215094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,256,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,256,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,256,0.009022222624884712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,64,0.001810666587617662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,128,0.004991111242108875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,128,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,32,0.0018088887963030075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,65536,0.013048888908492194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,256,128,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,256,1024,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,64,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,51200,0.011744000017642975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,32,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,256,768,0.005481777919663324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,65536,0.04657422171698677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,65536,0.04897155695491367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,51200,0.03760177890459696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,51200,0.038727111286587186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,12288,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,16384,0.01588355501492818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,16384,0.01588355501492818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,16384,0.04073244333267212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,10240,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,12288,0.01308444473478529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,12288,0.013273778061072031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,256,1536,0.0035582222044467926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,8192,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,12288,0.03291911217901442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,10240,0.012079111403889127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,10240,0.011818666425016193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,7168,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,10240,0.02812533246146308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,8192,0.010230222509966956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,8192,0.025053333905008104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,6144,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,7168,0.010022222167915767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,7168,0.009718221922715506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,7168,0.02264533274703556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,16384,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,5120,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,6144,0.009489778015348647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,65536,0.1367502212524414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,6144,0.00904622260067198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,4096,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,6144,0.02026844521363576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,5120,0.008917333351241218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,5120,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,3584,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,5120,0.018639999959203932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,4096,0.008221333225568136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,4096,0.00776533368561003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,3072,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,4096,0.01648622254530589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,3584,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,3584,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,3584,0.016019556257459853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,3072,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,3072,0.007131555842028723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,3072,0.01465777721669939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,2048,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,2560,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,2560,0.00665866666369968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,2560,0.013471999930010902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,1536,0.0035217776894569397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,2048,0.0069368887278768755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,2048,0.013108444710572561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,1536,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,1024,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,51200,0.10798311233520508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,1536,0.012113778127564324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,1024,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,768,0.002704888789190186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,1024,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,2560,0.005042666776312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,768,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,768,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,512,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,768,0.010061333576838175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,512,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,256,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,512,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,2048,0.006227555374304454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,512,0.009871111147933537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,128,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,256,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,256,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,1536,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,128,0.005542222410440445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,128,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,128,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,65536,0.0123848890264829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,1024,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,64,0.004956444518433677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,51200,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,32,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,16384,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,12288,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,10240,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,112,128,8192,0.010696888797812991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,65536,0.04766577813360426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,8192,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,7168,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,51200,0.038803554243511625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,16384,0.015596444408098856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,12288,0.013235555754767524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,6144,0.005160888863934411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,10240,0.011468444433477191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,4096,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,8192,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,3584,0.004955555415815777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,6144,0.008678221868144141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,5120,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,3072,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,2560,0.004954666727119022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,4096,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,112,128,256,0.009658666948477427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,2048,0.004580444345871608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,128,128,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,3584,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,1024,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,3072,0.007033777733643849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,768,0.002528000002106031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,2560,0.006322666588756774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,512,0.002451555596457587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,2048,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,1536,0.005983110931184556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,1024,0.005388444496525659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,128,0.0018808889306253856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,768,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,64,0.0018177777528762817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,512,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,32,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,128,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,65536,0.011758222348160215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,5120,0.004858666823969947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,64,0.005116444494989184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,51200,0.008343111309740279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,32,0.004585777719815572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,16384,0.006163555714819167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,7168,0.009646221995353699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,12288,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,10240,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,65536,0.047835555341508657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,8192,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,51200,0.03869422276814779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,16384,0.015670221712854173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,7168,0.005283555636803309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,1536,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,6144,0.004959110998445087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,12288,0.012764444781674279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,5120,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,10240,0.011703111231327057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,4096,0.004950222041871813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,3584,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,8192,0.010083555347389644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,7168,0.009513777991135916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,64,256,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,3072,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,5120,0.008043555749787224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,2560,0.00461422238085005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,4096,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,2048,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,1536,0.003567999849716822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,3584,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,1024,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,64,256,0.004874666531880696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,3072,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,2560,0.00647288892004225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,768,0.002528000002106031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,2048,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,256,0.0021262222694026097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,1024,0.005435555759403441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,128,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,768,0.005410666680998272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,512,0.005127111242877113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,64,0.0018053334206342697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,32,0.0017928888814316855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,256,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,128,0.004618666652176115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,64,0.004699555536111196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,32,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,6144,0.008732444710201686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,16384,0.371942228741116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,112,32,512,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,112,32,1536,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,16384,0.20323199696011016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,51200,0.6091306474473741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,51200,1.143202675713433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,12288,0.2796951134999593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,10240,0.20871378315819633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,16384,0.13352355692121717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,12288,0.1559777789645725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,12288,0.099254224035475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,10240,0.08386488755544026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,51200,0.3980702294243707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,8192,0.1106862227121989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,8192,0.1747199959225125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,10240,0.12654577361212835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,6144,0.12939377625783285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,5120,0.10911111036936443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,7168,0.09307289123535156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,6144,0.05964622232649061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,7168,0.06797689199447632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,4096,0.0914542211426629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,5120,0.04484355449676514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,5120,0.06632888979381986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,3584,0.07940355274412367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,4096,0.05541155404514737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,4096,0.03904977771970961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,3584,0.03452800048722161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,3584,0.04905600017971463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,2560,0.05765510929955376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,3072,0.03073599934577942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,3072,0.04270577761862013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,2048,0.048398223188188344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,2560,0.027127110295825537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,7168,0.1502808862262302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,2560,0.037222223149405584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,1536,0.036877334117889404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,2048,0.023482667075263128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,2048,0.0321066677570343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,1024,0.02711822258101569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,768,0.020469332734743755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,1536,0.018572444717089336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,1024,0.015199111567603217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,1536,0.02527644402450985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,512,0.01388177772363027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,768,0.013775111072593264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,6144,0.07879910866419475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,768,0.01959911154376136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,512,0.014301333162519666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,3072,0.06916711065504286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,128,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,256,0.011323555476135679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,64,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,128,0.011180444724029966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,32,0.0052800000541739995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,128,0.011237333218256632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,64,0.011751111182901593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,8192,0.08061422242058648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,32,0.015492444237073263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,1024,0.025253333979182776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,65536,256,0.008443555898136562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,65536,512,0.012477333347002665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,65536,256,0.01239377756913503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,65536,0.6292808850606283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,16384,0.2821235656738281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,65536,1.1023155848185222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,51200,0.47855202356974286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,12288,0.21207555135091147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,16384,0.16068266497717962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,16384,0.10753244161605835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,12288,0.12006399366590713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,65536,0.40909245279100204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,12288,0.0792728861172994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,51200,0.3105493386586507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,10240,0.10084799925486247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,10240,0.1940000057220459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,10240,0.06845600075191922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,8192,0.15909600257873535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,7168,0.13871910836961535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,8192,0.08607021967569987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,6144,0.11931110752953424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,8192,0.055481778250800244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,7168,0.049808889627456665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,51200,0.8672595553927951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,5120,0.10041866699854533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,6144,0.05063199996948242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,6144,0.0634711119863722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,4096,0.08170221911536323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,5120,0.037203556961483426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,5120,0.05347733365164863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,3584,0.07240088780721028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,3072,0.06300089094373915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,4096,0.0320995549360911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,4096,0.044454223579830594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,3584,0.02870755394299825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,2560,0.05287377701865303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,3072,0.025735111700163946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,2048,0.042782220575544566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,2560,0.02255200015174018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,2560,0.03046577837732103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,3584,0.039729777309629656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,3072,0.034958223501841225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,1536,0.03356533249219259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,1024,0.024136000209384497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,2048,0.019779556327395968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,2048,0.026825777358478967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,768,0.018230222993426852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,1536,0.01614488826857673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,1536,0.020954666866196528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,1024,0.01346933344999949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,1024,0.021262221866183813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,768,0.012292444705963135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,256,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,7168,0.07251555389828153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,128,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,512,0.012483555409643384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,128,0.010022222167915767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,64,0.0042568888101312844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,256,0.010710222025712332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,32,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,64,0.010031999813185798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,32,0.01311644415060679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,51200,512,0.013129777378506131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,65536,0.33395910263061523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,65536,0.22673511505126953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,51200,0.26073333952162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,65536,0.14074577225579157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,16384,0.08624977535671657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,256,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,51200,0.17415911621517607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,12288,0.06656177838643391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,51200,0.11189244853125678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,16384,0.060434665944841176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,16384,0.04276355438762241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,10240,0.05521066652403938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,12288,0.04615555538071526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,8192,0.04599644409285652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,12288,0.033201777272754245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,7168,0.03997688823276096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,10240,0.03975822197066413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,8192,0.023398222194777593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,8192,0.034834666384591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,6144,0.03467377689149644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,7168,0.020966221888860066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,7168,0.02979555394914415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,6144,0.020792000823550753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,768,0.016240888171725802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,4096,0.025027554896142747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,6144,0.026752889156341553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,5120,0.01696888936890496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,5120,0.02295911146534814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,3584,0.021723555194007024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,51200,128,0.010001777774757808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,4096,0.019314666589101154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,3072,0.019256000717480976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,3584,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,2560,0.016011555989583332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,3584,0.01753866672515869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,3072,0.012983111043771109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,2048,0.013201778133710226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,3072,0.015596444408098856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,2560,0.011890666352377998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,10240,0.02756711178355747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,1536,0.010327999790509542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,2560,0.013804444836245643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,2048,0.011044444309340583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,1024,0.008087111016114553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,2048,0.0124195557501581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,768,0.00657155571712388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,1536,0.009367999931176504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,1024,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,1536,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,512,0.005063111169470681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,5120,0.029751112063725788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,1024,0.010650667051474253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,768,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,51200,512,0.011040888726711273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,768,0.008914666871229807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,128,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,512,0.007122666471534305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,256,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,512,0.007096000015735626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,64,0.0026355555488003623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,4096,0.01514400045077006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,256,0.006588444527652528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,128,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,64,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,65536,0.2408391104804145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,16384,32,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,65536,0.17516444789038765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,51200,0.18695199489593506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,51200,0.1336711115307278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,65536,0.12450933456420898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,16384,0.06349688768386841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,12288,0.04891644583808052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,16384,0.037704888317320086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,256,0.0032888888898822996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,10240,0.04153866569201151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,12288,0.03014311194419861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,16384,32,0.0028524444335036804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,16384,128,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,12288,0.036321779092152916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,8192,0.035183111826578774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,10240,0.031762666172451444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,7168,0.02998222245110406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,10240,0.02516711089346144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,6144,0.026671111583709717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,8192,0.02774577836195628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,7168,0.02162222233083513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,7168,0.024093333217832778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,6144,0.019828443725903828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,5120,0.018370666437678866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,5120,0.016030222177505493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,4096,0.01998488936159346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,51200,0.10110933250851101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,4096,0.014574221438831754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,16384,0.048375109831492104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,3584,0.016678222351604037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,4096,0.015601777368121676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,3584,0.013359111216333179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,2560,0.012807110945383707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,3584,0.014290667242474027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,3072,0.013081777426931592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,2048,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,8192,0.022444445225927565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,2560,0.011439111497667102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,1536,0.008979555633332994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,2560,0.01204355557759603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,2048,0.009786666267448002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,5120,0.022945778237448797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,2048,0.011028444601429833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,1536,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,6144,0.021350221501456365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,768,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,1536,0.009710222482681274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,1024,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,512,0.004206222378545337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,768,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,1024,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,256,0.003846222327815162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,768,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,512,0.006948444578382704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,128,0.0032186667538351486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,512,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,64,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,256,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,32,0.0029884444342719186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,256,0.00600355573826366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,3072,0.014867555763986377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,128,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,3072,0.012473777764373355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,64,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,12288,32,0.0070053330726093715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,65536,0.2141937812169393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,65536,0.16632889376746282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,51200,0.16804088486565483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,16384,0.05857333209779528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,65536,0.12216356065538193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,12288,0.04616444309552511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,12288,1024,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,51200,0.09741066561804877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,16384,0.0355697770913442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,10240,0.03803822067048814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,12288,0.028352889749738906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,8192,0.03284710976812575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,12288,0.035000890493392944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,10240,0.02404088940885332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,7168,0.027691556347741023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,10240,0.030457778109444514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,8192,0.021361778179804485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,6144,0.024271999796231587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,8192,0.02606933315594991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,7168,0.01900622248649597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,5120,0.020767110917303298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,7168,0.023004444109068975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,6144,0.02017866737312741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,4096,0.017723555366198223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,5120,0.015504888362354703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,5120,0.017602667212486267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,3584,0.013992889059914483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,4096,0.014118222726715935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,4096,0.015173332558737861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,3072,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,51200,0.127527117729187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,3584,0.014076444837782117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,2560,0.010848888920413123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,16384,0.045797334776984326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,3072,0.012134222520722283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,2048,0.009648000200589498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,3072,0.012894222305880653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,2560,0.010820444259378644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,1536,0.007962666451931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,2560,0.011731555892361535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,2048,0.00925333301226298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,1024,0.00630044440428416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,2048,0.010582221878899468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,1536,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,12288,128,0.006676444576846228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,6144,0.017416889468828838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,768,0.0047377778424157035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,512,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,1024,0.007592000067234039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,768,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,256,0.003149333306484752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,1024,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,512,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,128,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,512,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,256,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,64,0.002493333278430833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,256,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,128,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,10240,3584,0.01313777764638265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,10240,32,0.002830222249031067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,128,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,64,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,65536,0.18340888288285997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,32,0.0069919998447100324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,65536,0.12241510550181071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,51200,0.13920177353752985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,16384,0.04948000113169352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,51200,0.05769155422846476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,1536,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,51200,0.09496621953116523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,12288,0.03828000028928121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,16384,0.02314044369591607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,10240,0.03208355440033807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,16384,0.034458667039871216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,8192,0.0270471109284295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,10240,768,0.007061333292060428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,10240,0.01629955569903056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,7168,0.023565333750512864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,10240,0.02342133389578925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,8192,0.014493332968817817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,6144,0.020399999287393358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,8192,0.02001066671477424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,7168,0.013474666410022311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,5120,0.017516444126764934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,7168,0.017817777064111497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,6144,0.012436444560686747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,4096,0.015135111080275642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,5120,0.011513777905040316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,3584,0.013072888884279462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,5120,0.014295111099878946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,4096,0.010669333239396414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,3072,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,4096,0.01274577776590983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,65536,0.07062311304940118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,3584,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,2560,0.009988444546858469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,3584,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,3072,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,2048,0.00867022242810991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,3072,0.010875555376211802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,2560,0.008644444247086843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,1536,0.007340444458855524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,2048,0.007438222567240397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,2560,0.010028444230556488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,1024,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,2048,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,1536,0.0070186663005087115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,768,0.004570666700601578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,12288,0.027270222703615826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,512,0.0037342223028341928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,1024,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,768,0.006325333482689328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,256,0.0029528888149393927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,768,0.006325333482689328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,512,0.006224888894293044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,128,0.002534222271707323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,512,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,256,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,6144,0.016023111012246873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,64,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,128,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,8192,32,0.002491555487116178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,128,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,65536,0.162063995997111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,64,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,32,0.006224000205596288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,51200,0.12962578402625188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,65536,0.1224035554462009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,16384,0.04615111152331034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,12288,0.018734221657117207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,51200,0.0927128858036465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,8192,1536,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,8192,1024,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,16384,0.022283555732833013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,51200,0.055826663970947266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,16384,0.03409511182043288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,10240,0.03052088949415419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,12288,0.017901332841979135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,8192,0.02608088817861345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,12288,0.026012443833880957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,7168,0.022652443912294176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,10240,0.015757333901193406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,10240,0.022661333282788593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,8192,0.014087999860445658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,8192,0.019726221760114033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,7168,0.013088888592190213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,6144,0.019567999574873183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,5120,0.016837333639462788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,7168,0.01722133325205909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,6144,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,4096,0.014476444986131458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,5120,0.011069333387745751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,6144,0.01579466627703773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,3584,0.011413333316644033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,5120,0.013953778478834363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,4096,0.00964355551534229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,3072,0.010984888507260216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,4096,0.01255555533700519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,3584,0.008902221918106079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,65536,0.06846311357286242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,3072,0.008055999875068665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,3584,0.011533333195580376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,2048,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,3072,0.010679999987284342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,12288,0.03647999962170919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,1536,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,1024,0.005295111073387994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,2048,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,2560,0.009941333697901832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,768,0.00453599997692638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,1536,0.006912888752089606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,1024,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,1536,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,512,0.003846222327815162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,768,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,1024,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,768,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,512,0.005988444305128521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,128,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,512,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,256,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,64,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,128,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,32,0.0024906665914588505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,128,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,64,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,65536,0.14175466696421304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,32,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,7168,2560,0.007689778175618913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,51200,0.1126835611131456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,2048,0.00847822262181176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,65536,0.06625422504213122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,16384,0.04229955540763008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,51200,0.08806844552357991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,51200,0.05362488826115926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,256,0.0031644445326593188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,12288,0.03492800063557095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,16384,0.021369778447681006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,16384,0.03253777821858724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,12288,0.017306667235162523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,8192,0.020842666427294414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,12288,0.024708444873491924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,10240,0.01540177729394701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,7168,0.018100443813535903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,10240,0.02162488963868883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,8192,0.01367022262679206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,6144,0.015343111422326831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,8192,0.018961777289708454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,7168,2560,0.009158222211731805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,7168,0.012764444781674279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,5120,0.013440888788965015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,7168,0.016757332616382174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,6144,0.011789333489206104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,4096,0.01237155579858356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,6144,0.015275556180212231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,3584,0.010285333626800114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,5120,0.013799111048380533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,4096,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,65536,0.11299467086791992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,3072,0.009726222190592024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,4096,0.012254222399658628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,3584,0.008815999660227034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,3584,0.011354666617181567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,3072,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,10240,0.029827555020650227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,2048,0.0075839997993575195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,2560,0.007595555649863348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,2560,0.00905688852071762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,7168,256,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,1536,0.005969777703285217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,2048,0.007240888973077138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,1024,0.004781333522664176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,2048,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,1536,0.0069288888739215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,768,0.00388711111413108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,1536,0.007213333414660559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,1024,0.006395555618736479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,512,0.003532444437344869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,1024,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,768,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,768,0.006303999986913469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,5120,0.010644444160991244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,512,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,128,0.0024622221373849446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,512,0.005736888696750005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,256,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,64,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,256,0.005533333453867171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,6144,128,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,2560,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,32,0.002191999927163124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,128,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,3072,0.0105253333846728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,65536,0.11408978038363987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,64,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,6144,32,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,65536,0.06376533375846015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,16384,0.03439377744992574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,65536,0.10942222012413873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,12288,0.028400888045628864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,51200,0.0514373348818885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,10240,0.025073778298166063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,51200,0.08492533365885417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,16384,0.020261334048377145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,12288,0.016859556237856548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,16384,0.031207998593648274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,12288,0.023853333459960088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,10240,0.014936889211336771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,7168,0.015385778413878547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,10240,0.020991999242040846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,6144,256,0.0028320000403457214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,6144,0.01367288910680347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,8192,0.018445332845052082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,7168,0.01236088905069563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,5120,0.011891555454995898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,7168,0.01651822196112739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,6144,0.011184000306659274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,4096,0.01014311114947001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,5120,0.010017777482668558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,6144,0.01483022173245748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,3584,0.01001422190003925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,4096,0.008667555948098501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,5120,0.013407111167907715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,3072,0.009324444664849175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,4096,0.01186222251918581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,2560,0.008486222061845992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,3072,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,51200,0.09066933393478394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,2048,0.00737955586777793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,3072,0.009684444301658207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,2560,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,1536,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,2048,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,8192,0.021969777014520433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,2048,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,1024,0.004734222259786394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,1536,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,768,0.00388711111413108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,1536,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,8192,0.013368888861603208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,1024,0.006490666833188798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,512,0.0034888887570963967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,1024,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,768,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,256,0.0028106667515304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,512,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,128,0.002536000063021978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,768,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,512,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,64,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,128,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,256,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,5120,32,0.002464888824356927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,3584,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,128,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,65536,0.09034222364425659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,64,0.005487111293607288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,3584,0.01092088884777493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,51200,0.073908441596561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,32,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,65536,0.06167022387186686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,16384,0.029139555162853662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,65536,0.0669413341416253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,5120,2560,0.008787555827034844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,12288,0.02458222210407257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,51200,0.049253334601720176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,51200,0.05272355675697327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,16384,0.01979822251531813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,10240,0.021293333835071985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,12288,0.016183111402723525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,8192,0.015285332997639975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,12288,0.016152888536453247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,10240,0.014459555347760519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,7168,0.013400889105266996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,8192,0.01298577752378252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,10240,0.014646222194035849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,6144,0.011791110866599612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,7168,0.01178311142656538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,8192,0.013506666653686099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,5120,0.010447999669445885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,7168,0.012072000238630505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,5120,256,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,4096,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,5120,0.009470221896966299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,6144,0.011011555790901184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,3584,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,5120,0.010024888647927178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,4096,0.00846400029129452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,3072,0.007371555599901412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,3584,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,4096,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,2560,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,3072,0.00795555528667238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,2048,0.00563466673096021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,3072,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,2560,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,1536,0.004534222185611725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,2560,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,2048,0.007181333170996771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,1024,0.0035457776652442086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,2048,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,16384,0.02092711130777995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,768,0.0031884445084465873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,1536,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,512,0.002814222127199173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,1024,0.006309333360857434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,768,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,256,0.0025804444319672054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,1024,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,768,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,512,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,128,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,512,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,256,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,6144,0.010614222122563256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,64,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,4096,32,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,128,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,256,0.005129777722888523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,128,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,65536,0.08628711435529922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,51200,0.06908710797627766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,64,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,65536,0.06073955694834391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,16384,0.027791112661361694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,3584,0.008319999608728621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,12288,0.02199911077817281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,51200,0.051233778397242226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,51200,0.04853511187765333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,16384,0.019521777828534443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,10240,0.01958933307064904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,16384,0.01995466649532318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,12288,0.015837333268589444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,8192,0.015365334020720588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,4096,1536,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,12288,0.015797333584891427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,7168,0.012882666455374824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,10240,0.014155555102560254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,8192,0.012470222181744046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,6144,0.011363555159833698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,7168,0.01131377783086565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,8192,0.012899555265903473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,5120,0.010040889183680216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,7168,0.011585777832402123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,6144,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,4096,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,5120,0.009100444614887238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,6144,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,3584,0.007928888830873702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,5120,0.009659555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,3072,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,4096,0.00851111114025116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,4096,0.00831911133395301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,3584,0.008182221816645728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,2560,0.006542222367392645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,4096,32,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,3584,0.007793777518802219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,65536,0.06622133652369182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,3072,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,2560,0.0074337778819931885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,1536,0.004576888763242298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,2560,0.0069439998931354946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,2048,0.0069884442620807225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,1024,0.0035146665241983202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,2048,0.006568000134494569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,1536,0.006657777975002925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,768,0.003177777760558658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,1536,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,10240,0.014402666025691561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,1024,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,1024,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,768,0.006049777898523543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,256,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,768,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,512,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,512,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,128,0.002463999928699599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,256,0.005672889037264719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,64,0.0021537778278191886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,256,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,128,0.005636444522274866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,32,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,128,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,65536,0.07531288597318861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,2048,0.005488889084921942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,64,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3584,32,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,51200,0.060841778914133705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,65536,0.059306667910681836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,16384,0.022936888866954382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,65536,0.06349155637953016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,12288,0.019377777973810833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,51200,0.047247111797332764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,51200,0.04914933443069458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,16384,0.019154666198624503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,10240,0.016301333904266357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,16384,0.019124444988038804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3584,512,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,8192,0.014518222875065274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,10240,0.01386044422785441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,12288,0.015327110886573792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,7168,0.013173333472675748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,10240,0.013753777576817406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,8192,0.01201333353916804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,6144,0.011958222422334882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,7168,0.011033777561452655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,5120,0.009736888938479954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,7168,0.011377777490350934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,6144,0.009776888622177972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,4096,0.008618666893906063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,6144,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,5120,0.009000889129108852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,3584,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,5120,0.009482666850090027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3584,3072,0.007735110819339752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,4096,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,3072,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,4096,0.008007999923494127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,3584,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,3584,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,3072,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,2048,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,3072,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,1536,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,2560,0.006673777682913675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,2048,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,1024,0.004211555752489301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,12288,0.015862221519152325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,2048,0.006521777974234686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,1536,0.006835555450783835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,768,0.0037644443412621817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,1536,0.00611200018061532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,1024,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,512,0.0031946665710873077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,1024,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,768,0.006181333214044571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,256,0.0028106667515304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,8192,0.012480888929631976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,128,0.002540444541308615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,512,0.005980444451173146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,512,0.005286222116814719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,256,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,64,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,128,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,32,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,256,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,128,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,65536,0.06471644507514106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,64,0.005013333426581489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,51200,0.05259822143448723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,3072,2560,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,32,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,16384,0.02051555613676707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,65536,0.05787822273042467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,3072,2560,0.007449777589903937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,51200,0.04615911179118686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,51200,0.04728088776270548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,16384,0.018571555614471436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,12288,0.015388444066047668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,10240,0.014799111419253878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,12288,0.015108444624476962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,8192,0.012952889005343119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,10240,0.013427555561065674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,10240,0.013391111459996967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,3072,768,0.005623111294375525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,8192,0.01106844428512785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,8192,0.012034667034943899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,6144,0.011040888726711273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,7168,0.010587555666764578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,7168,0.011039999624093374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,5120,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,6144,0.009848888549539778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,6144,0.010392000277837118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,4096,0.008596444295512306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,5120,0.009003555609120263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,12288,0.017190222938855488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,5120,0.008977777428097194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,3584,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,16384,0.0188711103465822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,4096,0.008381333616044786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,3072,0.006728000111050076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,4096,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,3584,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,3072,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,3584,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,2048,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,3072,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,2560,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,1536,0.004239999999602635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,2560,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,2048,0.007076444725195567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,1024,0.0035297779573334586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,2048,0.0063902222447925145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,1536,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,768,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,1536,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,1024,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,512,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,1024,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,768,0.006213333457708359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,256,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,768,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,512,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,65536,0.06089422437879774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,128,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,512,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,64,0.0021946666141351065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,256,0.005195555587609609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,32,0.0021964444054497615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2560,128,0.00563466673096021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,65536,0.05336177680227491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,128,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,2560,0.005769777629110549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,64,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2560,32,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,65536,0.05668266614278158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,16384,0.01755199995305803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,65536,0.0581351121266683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,12288,0.01349066694577535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,51200,0.04526222083303663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,51200,0.04537777768241035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2560,7168,0.012142221960756512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,10240,0.012298666768603854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,16384,0.01850133306450314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,8192,0.011040888726711273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,12288,0.014708444476127625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,16384,0.01795377830664317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,7168,0.010598222414652506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,12288,0.014879110786649915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,10240,0.01240711079703437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,6144,0.009827555881606208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,8192,0.011047999891969891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,10240,0.013167999684810638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,5120,0.0090506664580769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,8192,0.011553777588738335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,4096,0.00793599999613232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,6144,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,5120,0.008982222113344405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,3584,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,5120,0.008761777646011775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,4096,0.00849333322710461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,3072,0.0063440000845326324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,4096,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,3584,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,2560,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,3584,0.0075902218619982404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,3072,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,2048,0.004563555535342959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,3072,0.007324444750944774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,51200,0.043097777499092944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,2560,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,1536,0.0038542221817705366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,2560,0.00666133314371109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,2048,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,1024,0.0032133333798911837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,2048,0.00638488887084855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,1536,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,768,0.0028782222006056043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,1536,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,1024,0.006303999986913469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,512,0.0025333333760499954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,1024,0.005842666659090254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,768,0.005994666781690385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,7168,0.010330666270520952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,256,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,768,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,512,0.0058142224119769196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,128,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,7168,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,64,0.0021128888345426982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,6144,0.009503111243247986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,512,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,256,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,2048,32,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,2048,128,0.00562755556570159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,256,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,128,0.005087999833954705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,64,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,2048,32,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,65536,0.0556817783249749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,16384,0.014702222413486905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,65536,0.055326223373413086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,12288,0.012441777520709567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,51200,0.04471910993258158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,51200,0.04381422201792399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,16384,0.017694221602545846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,10240,0.011386666860845355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,16384,0.017788444956143696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,12288,0.013437333206335703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,8192,0.010552888943089379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,12288,0.014131555954615274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,10240,0.012416000167528788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,10240,0.012501333322789935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,8192,0.01091466678513421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,6144,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,8192,0.011122666299343109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,7168,0.01035822182893753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,7168,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,6144,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,4096,0.007753777835104201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,6144,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,65536,0.04109155469470554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,5120,0.008954666554927826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,3584,0.006426666759782367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,5120,0.008563555777072906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,4096,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,51200,0.03260711166593764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,3072,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,2560,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,4096,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,3584,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,2048,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,3072,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,3584,0.007400888535711501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,1536,0.004179555508825514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,2560,0.007341333561473423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,3072,0.0069884442620807225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,1024,0.003495111233658261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,2048,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,2560,0.006674666785531574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,768,0.0031786666562159858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,1536,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,2048,0.006440889090299606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,512,0.0028444443725877335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,1024,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,1536,0.00599644457300504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,7168,0.009881777895821465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,256,0.0024702221983008915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,768,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,1024,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,128,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,512,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,768,0.005623111294375525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,64,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,32,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,512,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,128,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,65536,0.030326220724317763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,256,0.004946666873163647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1536,5120,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,128,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,51200,0.024493333366182115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,64,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,16384,0.011733333269755045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1536,32,0.0052044445441828836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,65536,0.0547964440451728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,12288,0.012712000144852532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,51200,0.044136001004113086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,65536,0.05271555648909675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,10240,0.011413333316644033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,16384,0.016202666693263583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,51200,0.04219555523660448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,8192,0.010275555153687796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,12288,0.013390222357379066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,16384,0.017247110605239868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,7168,0.009338666995366415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,10240,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,12288,0.013330666555298699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,6144,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,10240,0.011918221910794577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,8192,0.010467555787828235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,5120,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,7168,0.010052444206343757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,6144,0.009333333207501305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,4096,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,5120,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,5120,0.008407110969225565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,3584,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1536,256,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,4096,0.008361777497662438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,3072,0.005649777750174205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,4096,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,3584,0.00792622235086229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,2560,0.005024000174469418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,3584,0.007338666253619724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,3072,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,2048,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,3072,0.006984000404675801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,2560,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,1536,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,2560,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,2048,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,1024,0.0036675557494163513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,2048,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,8192,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,1536,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,768,0.0031964443624019623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,7168,0.010055999788973067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,1536,0.005962666538026597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,1024,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,512,0.0028808888875775864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,256,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,768,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,512,0.0057751110030545135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,768,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,128,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,256,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,512,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,64,0.0019706665641731685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,256,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,1024,32,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,128,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,65536,0.023583999938435022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,128,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,51200,0.020469332734743755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,64,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,32,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,65536,0.054372443093193897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,16384,0.012666666673289405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,65536,0.05156177613470289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,12288,0.010695999695195092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,51200,0.043868445687823825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,51200,0.0417173339260949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,16384,0.01587022178702884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,10240,0.010058666268984476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,16384,0.01621955633163452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,12288,0.01346577786737018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,8192,0.008935111264387766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,10240,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,7168,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,8192,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,6144,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,1024,6144,0.009472000102202097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,8192,0.010528888967302112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,1024,1024,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,7168,0.010341333018408881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,5120,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,4096,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,7168,0.009710222482681274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,6144,0.009340444372759925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,5120,0.008940444224410588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,3584,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,6144,0.009093333449628618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,4096,0.008267555799749162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,3072,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,5120,0.008482666479216682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,2560,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,4096,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,3072,0.007329777710967594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,3584,0.0074879998962084455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,2048,0.00452888881166776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,3072,0.0069875551594628235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,1536,0.0038311113086011675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,2560,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,2048,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,1024,0.0031724443866146933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,2048,0.006314666734801398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,1536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,768,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,1536,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,512,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,12288,0.013081777426931592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,1024,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,768,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,10240,0.011712888876597086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,512,0.005758222192525864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,128,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,256,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,512,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,256,0.005153777698675792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,768,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,128,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,65536,0.018745778335465323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,128,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,3584,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,64,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,51200,0.015410666664441427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,32,0.005015999906592899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,65536,0.05317155520121256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,16384,0.010054222411579555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,768,2560,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,65536,0.050718221399519176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,12288,0.008638222184446123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,51200,0.039510223600599505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,10240,0.007794666621420119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,16384,0.015862221519152325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,51200,0.041249778535630964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,8192,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,12288,0.013421333498424955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,16384,0.015930665863884818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,7168,0.008568888737095727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,10240,0.012040889097584618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,12288,0.012846222354306115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,10240,0.011701333026091257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,8192,0.010818666881985135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,7168,0.010047111246320937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,8192,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,5120,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,7168,0.009673777553770278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,6144,0.009679111341635386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,768,768,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,6144,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,5120,0.008803555534945594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,3584,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,5120,0.008548444343937768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,4096,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,3072,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,4096,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,3584,0.00775733341773351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,2560,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,3584,0.007496889266702864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,3072,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,2048,0.00424177779091729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,1536,0.004042666819360522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,2560,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,2048,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,2560,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,1024,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,1536,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,2048,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,768,0.002838222309947014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,1536,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,1024,0.006225777582989798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,512,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,768,0.006107555495368109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,1024,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,256,0.0021991110924217436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,768,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,6144,0.00906755526860555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,128,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,512,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,256,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,64,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,32,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,128,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,512,4096,0.006882666713661618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,65536,0.015277332729763456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,128,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,51200,0.013176889055305056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,64,0.005386666705211003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,65536,0.04714488983154297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,32,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,16384,0.008922666311264038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,51200,0.037662221325768366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,51200,0.03927910990185208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,12288,0.007330666813585494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,3072,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,10240,0.006785777707894643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,16384,0.015791111522250705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,12288,0.013406222065289816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,8192,0.00867911097076204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,10240,0.01204177737236023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,12288,0.012740444805887012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,7168,0.008014221986134848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,10240,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,8192,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,512,512,0.00564444437623024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,6144,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,7168,0.010071111222108206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,5120,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,6144,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,512,256,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,4096,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,6144,0.008978666530715095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,5120,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,5120,0.008431110945012834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,3584,0.005958222266700532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,65536,0.049568888213899404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,4096,0.0075911109646161394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,3072,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,3584,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,3584,0.007365333537260692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,2560,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,3072,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,3072,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,2048,0.004217777815130022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,2560,0.007197333706749811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,2560,0.00664444433318244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,1536,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,8192,0.0103502223889033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,2048,0.0069919998447100324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,2048,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,7168,0.00960000024901496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,1536,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,768,0.0028239999794297745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,1536,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,1024,0.006127111199829314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,512,0.002453333387772242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,1024,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,768,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,256,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,4096,0.008269333177142674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,512,0.005772444523043103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,128,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,512,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,256,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,16384,0.016146666473812528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,64,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,32,0.0018577778504954444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,256,128,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,65536,0.013408000270525614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,128,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,64,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,32,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,65536,0.04722311099370321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,256,1024,0.002967111145456632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,16384,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,51200,0.037680887513690524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,51200,0.038033776813083224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,12288,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,16384,0.015841777125994366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,10240,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,16384,0.0155395550860299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,12288,0.013129777378506131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,12288,0.012780444489585029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,8192,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,10240,0.012094222009181976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,7168,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,10240,0.011418666276666852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,8192,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,6144,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,256,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,8192,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,7168,0.010064889159467485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,5120,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,6144,0.009697777529557547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,7168,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,4096,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,5120,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,51200,0.011398221883508893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,3584,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,5120,0.008278222547637092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,4096,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,3072,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,4096,0.007603555917739868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,2560,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,3584,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,3072,0.0073724447025193115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,2048,0.0041999999019834726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,3072,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,2560,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,1536,0.0035546666218174826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,2560,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,2048,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,1024,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,2048,0.006128888991143968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,1536,0.006656000183688269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,768,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,1536,0.005966222120655908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,1024,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,512,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,1024,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,768,0.006187555690606435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,256,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,768,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,512,0.005778666585683823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,128,0.0018702221827374564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,512,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,65536,0.046282665597067944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,256,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,64,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,256,0.005196444276306364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,128,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,97,128,3584,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,128,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,65536,0.011639110743999481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,64,0.0047422221137417685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,51200,0.009640889035330879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,16384,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,32,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,12288,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,65536,0.047241777181625366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,256,768,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,51200,0.03803733322355483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,10240,0.005647999958859549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,8192,0.0053137776752312975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,16384,0.015503111812803479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,7168,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,6144,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,10240,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,8192,0.010424888796276517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,5120,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,7168,0.009400889277458191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,4096,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,6144,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,128,6144,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,3072,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,128,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,5120,0.0083155557513237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,4096,0.007340444458855524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,2560,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,2048,0.004577777865860197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,3584,0.0070168889231152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,1536,0.0034933334423436057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,1024,0.0028355556229750314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,3072,0.006748444504208035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,768,0.002550222186578645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,2560,0.006660444455014334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,2048,0.0061919999619325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,512,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,256,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,1536,0.005715555614895291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,128,0.0018862223045693503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,1024,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,64,0.0017742222795883815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,768,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,32,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,12288,0.01278222186697854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,256,0.005314666777849197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,65536,0.01241422196229299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,128,0.004584889031118817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,51200,0.010270222193664974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,64,0.004589333302444882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,16384,0.008319999608728621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,12288,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,32,0.004569777597983678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,10240,0.005642666584915585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,65536,0.047278222110536366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,64,3584,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,8192,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,51200,0.03769955701298184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,7168,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,16384,0.015452444553375244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,6144,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,12288,0.012780444489585029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,5120,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,4096,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,10240,0.011391110718250275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,8192,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,7168,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,3072,0.004965333475006951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,6144,0.00865955568022198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,2560,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,5120,0.008351999852392409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,2048,0.004538666870858935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,1536,0.003552000141806073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,3584,0.007328888608349695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,3072,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,1024,0.002838222309947014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,2560,0.0063031112982167145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,768,0.0025351111673646504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,2048,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,512,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,256,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,1536,0.005766222046481238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,64,512,0.005046222358942032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,128,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,64,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,1024,0.005320888840489917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,768,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,32,0.0018257778137922287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,512,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,256,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,128,0.004874666531880696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,64,0.004588444613748127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,32,0.004594666676388847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,97,32,3584,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,16384,0.37291733423868817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,97,32,4096,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,51200,0.6052311261494955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,12288,0.27552888128492564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,51200,1.1510124206542969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,16384,0.13344266679551867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,12288,0.14923289087083605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,10240,0.20711110697852242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,51200,0.3980337778727214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,12288,0.10382933086819118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,16384,0.20303022861480713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,10240,0.12629955344729954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,8192,0.17217689090304902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,10240,0.0841368900405036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,16384,0.29085156652662486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,8192,0.1067848867840237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,12288,0.2108915514416165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,7168,0.15027199851142034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,8192,0.06843555635876127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,7168,0.09028444687525432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,10240,0.18142578336927626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,7168,0.06653866502973768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,5120,0.10973155498504639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,6144,0.07885155412885878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,8192,0.14801866478390163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,6144,0.05570489168167114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,7168,0.1267484426498413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,4096,0.09162844551934136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,5120,0.06783377461963229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,5120,0.048804445399178394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,3584,0.07858844598134358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,6144,0.11235822571648492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,4096,0.054663111766179405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,4096,0.04090222385194566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,3072,0.06892089049021403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,5120,0.09576533238093059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,3584,0.03494311041302151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,3584,0.04896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,6144,0.12886755996280247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,2560,0.057578669653998486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,3072,0.042678223715888135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,3072,0.031341334184010826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,2048,0.04628444380230374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,2560,0.02937777837117513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,2560,0.03728355632887946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,3072,0.06457422176996867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,2560,0.05713511175579495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,1536,0.03656889001528422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,2048,0.032264888286590576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,2048,0.023619555764728125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,1024,0.026996443669001263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,2048,0.048559112681282886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,1536,0.019126221537590027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,1536,0.025408888856569927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,768,0.02032088902261522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,1536,0.04115822248988681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,1024,0.025801777839660645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,1024,0.015486222174432544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,512,0.01367200000418557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,768,0.013772444592581855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,768,0.021194666624069214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,4096,0.0794888867272271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,1024,0.03427911135885451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,256,0.008286221987671321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,768,0.03007555670208401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,3584,0.0723031096988254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,512,0.012545777691735161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,512,0.015319110618697273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,128,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,256,0.011275555524561139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,512,0.025607110725508794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,256,0.012375999655988483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,64,0.0046364445653226645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,65536,32,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,128,0.011376000112957425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,128,0.01622577839427524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,64,0.01145333300034205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,51200,0.9135697682698568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,65536,32,0.015472888946533203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,65536,0.6280870967441136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,51200,0.4763840039571126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,65536,1.0934311548868816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,128,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,256,0.02053511142730713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,51200,0.8615733252631294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,16384,0.16034932931264242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,16384,0.277183108859592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,65536,0.4095475673675537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,16384,0.11066755321290757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,12288,0.20965422524346244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,12288,0.11975822183820937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,51200,0.32200532489352757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,10240,0.1708408859041002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,10240,0.10138755374484593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,8192,0.1383662223815918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,10240,0.06885333193673028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,16384,0.2345066600375705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,12288,0.17617777983347574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,8192,0.08561688661575317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,7168,0.12080177995893691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,10240,0.148016889890035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,8192,0.06080977784262764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,6144,0.10476177930831909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,7168,0.07241066959169176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,7168,0.050269332196977407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,12288,0.08846933311886257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,8192,0.12025245030721028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,6144,0.06347911225424872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,5120,0.08868266476525201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,7168,0.10733333561155532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,6144,0.04547199938032362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,6144,0.09401866462495591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,4096,0.07320266962051392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,5120,0.04039466712209914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,5120,0.053379555543263756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,3584,0.06312621964348687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,4096,0.044491555955674916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,4096,0.03432710965474447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,5120,0.08168177472220527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,3072,0.05518666572040982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,3584,0.029142220815022785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,4096,0.06851289007398817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,3584,0.06195111407174004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,3072,0.02714666724205017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,3072,0.03491200009981791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,2560,0.0463013317849901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,3072,0.055861334005991616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,2560,0.024671110841963027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,2048,0.037985778517193265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,2560,0.030652443567911785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,2560,0.04910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,2048,0.021199999584092036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,2048,0.02667466633849674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,2048,0.04335644510057238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,51200,0.7280613581339518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,1024,0.022145777940750122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,1536,0.016470222009552848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,1536,0.02113866640461816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,1536,0.03607022100024753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,3584,0.03975911272896661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,768,0.015949333707491558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,1024,0.013746666411558786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,1024,0.02342844506104787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,65536,0.9157199859619141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,512,0.011360000405046674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,1024,0.030437333716286555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,768,0.017078222499953378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,256,0.007608888877762689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,768,0.02552000019285414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,512,0.010864888628323873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,512,0.014544000228246054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,512,0.022050667140218947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,128,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,256,0.01001422190003925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,256,0.010495111346244812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,64,0.004063111212518481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,32,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,51200,1536,0.029620445436901514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,128,0.010098666780524785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,128,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,128,0.014408888088332282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,64,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,51200,32,0.013081777426931592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,65536,0.34131643507215714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,65536,0.22795110278659395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,51200,0.2676577832963732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,51200,0.1738204426235623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,65536,0.1466791099972195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,16384,0.08658311102125381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,51200,0.117667555809021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,16384,0.042555554045571216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,12288,0.06951644685533312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,16384,0.06049866808785332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,51200,256,0.018545778261290658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,16384,0.0601102246178521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,10240,0.057040890057881675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,12288,0.045906666252348155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,12288,0.04667288727230496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,10240,0.02872177627351549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,8192,0.049584001302719116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,10240,0.04136177897453308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,10240,0.039834668238957725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,8192,0.03479111194610596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,51200,768,0.012444444828563266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,7168,0.04148177636994256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,51200,0.15935199790530735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,65536,0.20317955811818442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,6144,0.03604266709751553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,8192,0.03749422232309977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,7168,0.021757332815064326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,5120,0.031088001198238794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,7168,0.02981511089536879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,6144,0.02050044470363193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,6144,0.026494221554862127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,7168,0.032403555181291364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,6144,0.029996444781621296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,5120,0.018214222457673814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,5120,0.02294755478700002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,12288,0.03489511211713155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,4096,0.0158924443854226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,5120,0.026529777381155226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,4096,0.0195448895295461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,3072,0.01981244484583537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,4096,0.02346666653951009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,3584,0.014319111903508505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,3584,0.017498667041460674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,3584,0.02183377742767334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,2560,0.016556445095274184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,3072,0.015467555986510383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,3072,0.013408888545301227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,8192,0.024716445141368445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,3072,0.019256000717480976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,2560,0.012279999752839407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,2560,0.013954665925767688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,2560,0.01798577772246467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,1536,0.011015111373530494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,2048,0.012479110724396177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,2048,0.01108977778090371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,4096,0.026542221506436665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,2048,0.017049777838918898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,1536,0.009375111096435124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,1536,0.010673777924643623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,3584,0.022641777992248535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,768,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,1536,0.015441776977645027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,1024,0.010784889260927836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,1024,0.01382844481203291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,512,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,768,0.009010666774378883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,768,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,768,0.012940444052219391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,256,0.0036071112586392295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,512,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,2048,0.013902222116788229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,512,0.010710222025712332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,256,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,128,0.0031893334041039148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,256,0.006820444431569841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,256,0.010053333308961656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,64,0.002806222273243798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,128,0.00666222224632899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,32,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,128,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,128,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,16384,1024,0.008766222331258986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,64,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,32,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,65536,0.27330043580796987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,1024,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,65536,0.17343910535176596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,51200,0.21089688936869302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,16384,0.06968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,51200,0.133424891365899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,65536,0.12862932682037354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,12288,0.05504444572660658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,51200,0.10156710942586263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,16384,512,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,16384,0.04844889044761658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,16384,0.0377768874168396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,10240,0.04494844542609321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,12288,0.029537777105967205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,16384,0.05176266696718004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,12288,0.03660533494419522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,8192,0.03769777880774604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,12288,0.04177422324816386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,10240,0.031558222240871854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,10240,0.02574133376280467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,7168,0.03299644589424133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,10240,0.0368755559126536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,8192,0.02199466692076789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,8192,0.02734311090575324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,51200,0.1328293350007799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,6144,0.028676443629794653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,65536,0.16472000545925564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,8192,0.0321760012043847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,7168,0.02405511173937056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,7168,0.019973332683245342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,5120,0.024707555770874023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,7168,0.02913244565327962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,6144,0.01904266741540697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,6144,0.02125866711139679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,5120,0.016893333858913846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,3584,0.018241778016090393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,5120,0.01812355551454756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,6144,0.026775999201668635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,4096,0.01482222146458096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,5120,0.023726221587922838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,3072,0.01607644392384423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,3584,0.013624889155228933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,3584,0.01442399952146742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,2560,0.013398222625255585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,3072,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,3584,0.019672888848516677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,3072,0.018245332770877414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,2048,0.011348444554540845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,2560,0.011791999969217511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,2560,0.012034667034943899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,2560,0.016945777667893302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,1536,0.009321778184837764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,2048,0.010630221830474006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,2048,0.016352888610627916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,4096,0.021104888783560857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,1024,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,1536,0.009633777870072259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,1536,0.008781332936551835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,1536,0.014868444866604276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,1024,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,768,0.006340444501903322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,1024,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,1024,0.01239466667175293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,3072,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,768,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,512,0.004563555535342959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,768,0.007703111403518253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,768,0.011756444142924415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,256,0.003528888854715559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,512,0.006908444480763541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,512,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,512,0.010105777945783403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,2048,0.011128000087208219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,128,0.0028799999919202593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,256,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,256,0.006035555568006303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,64,0.002510222295920054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,128,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,12288,32,0.003006222140457895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,128,0.007644444704055786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,64,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,4096,0.015492444237073263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,32,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,65536,0.24201777246263292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,65536,0.16614754994710287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,51200,0.19196800390879312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,65536,0.1240417824851142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,16384,0.06023910972807142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,51200,0.12764533360799155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,51200,0.09817688994937473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,16384,0.0454693337281545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,256,0.00998222248421775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,12288,0.04691911074850294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,12288,128,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,16384,0.0366888874106937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,10240,0.03821244504716661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,12288,0.029189334975348577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,12288,0.03485066692034403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,4096,0.021528000632921856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,12288,0.040271110004848905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,51200,0.12536799907684326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,10240,0.024515555964575872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,10240,0.030421333180533514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,7168,0.028001778655582007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,8192,0.021169778373506334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,8192,0.026329777306980554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,8192,0.03052266769938999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,6144,0.024158222807778254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,7168,0.019984000258975558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,7168,0.028569777806599934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,6144,0.02029333346419864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,5120,0.020784889658292133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,6144,0.017842666970358956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,65536,0.15659733613332114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,16384,0.05006844467586941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,4096,0.017849778135617573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,5120,0.017549332645204332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,5120,0.015855110353893705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,8192,0.03303022185961405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,3584,0.014099554883109199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,5120,0.022778666681713525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,4096,0.014088888963063559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,4096,0.015201777219772339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,3072,0.012432888978057437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,10240,0.0351200004418691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,4096,0.020229332976871066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,3584,0.01408266690042284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,3584,0.013099555340078143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,2560,0.011333333121405708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,3584,0.01958755486541324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,7168,0.022896000080638464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,3072,0.012824888858530255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,2048,0.009415999882751042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,2560,0.010827555424637265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,2560,0.011714666253990598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,3072,0.017910222212473553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,1536,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,2560,0.01657511128319634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,2048,0.009090666969617208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,2048,0.010589333044158088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,6144,0.025545777546034917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,2048,0.015789333317014907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,1536,0.00833511104186376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,1536,0.008966222405433655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,768,0.0047004446387290955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,1536,0.01402666668097178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,1024,0.007660444411966536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,512,0.003908444609906939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,1024,0.012077333198653327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,768,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,768,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,256,0.0031902222997612427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,768,0.011341333389282227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,512,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,512,0.0063271112740039825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,128,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,512,0.01038755559258991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,256,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,256,0.0063271112740039825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,256,0.009645333720578088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,64,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,128,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,128,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,32,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,128,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,64,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,10240,1024,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,65536,0.17665599452124703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,32,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,51200,0.13836177190144858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,65536,0.1238551139831543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,10240,1024,0.00831733312871721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,65536,0.07071644730038114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,16384,0.05197599861356947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,51200,0.05785422192679512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,16384,0.0351875556839837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,12288,0.03700177868207296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,16384,0.022999111149046157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,16384,0.045922666788101196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,3072,0.012096888489193387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,10240,0.030858665704727173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,65536,0.1408311128616333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,12288,0.018564444449212816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,12288,0.02702933384312524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,51200,0.11261155870225693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,8192,0.025447999437650044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,12288,0.03745066788461473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,10240,0.016302223006884258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,10240,0.023652444283167522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,7168,0.022982221510675218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,8192,0.014464888307783337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,10240,0.03353688783115811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,8192,0.020266667008399963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,6144,0.019878221882714164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,8192,0.029139555162853662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,7168,0.013544888959990608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,7168,0.0177831103404363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,51200,0.09439910782708062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,7168,0.02624711063173082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,6144,0.012438221938080259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,6144,0.01586844523747762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,4096,0.014136888914638095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,6144,0.02386844489309523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,5120,0.011389333340856763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,5120,0.01421688828203413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,3584,0.012192000117566852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,4096,0.010615111225181155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,5120,0.021893334057595994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,4096,0.012704888979593912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,3072,0.010793777803579966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,4096,0.01981599960062239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,3584,0.009633777870072259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,3584,0.011737777955002256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,2560,0.009637333452701569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,3584,0.018795554836591084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,2048,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,3072,0.010944888823562197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,2560,0.008077333370844523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,3072,0.01755466726091173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,2560,0.009979555176364051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,1536,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,2560,0.016390222642156813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,2048,0.00905600024594201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,1024,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,2048,0.015138667490747241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,1536,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,1536,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,1536,0.013639111485746173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,768,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,1024,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,1024,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,512,0.003489777859714296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,1024,0.011713777979214987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,768,0.006302222195598815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,768,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,256,0.00282222218811512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,768,0.011065777805116443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,512,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,512,0.005985777825117111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,128,0.002500444443689452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,512,0.010327999790509542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,3072,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,256,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,64,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,32,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,256,0.009704888694816166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,128,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,128,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,128,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,64,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,65536,0.1597999996609158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,8192,32,0.006018666757477655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,8192,5120,0.01717333330048455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,65536,0.12055022186703152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,65536,0.06849600209130181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,16384,0.04756888747215271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,51200,0.09264977773030598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,51200,0.05566755268308851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,16384,0.021893334057595994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,16384,0.03386844529045953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,12288,0.034416000048319496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,16384,0.04475111100408766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,65536,0.13973511589898005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,10240,0.02864977717399597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,12288,0.026022222306993272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,12288,0.017872888180944655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,12288,0.036884443627463445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,8192,0.02460355559984843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,10240,0.022359111242824133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,10240,0.015789333317014907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,2048,0.007678222325113084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,7168,0.02127200033929613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,10240,0.03267644511328803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,8192,0.014093332820468478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,51200,0.12933954927656385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,8192,0.019369777705934312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,6144,0.018553778529167175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,8192,0.02871911062134637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,7168,0.017343999611006845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,5120,0.01586844523747762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,7168,0.026011554731263056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,6144,0.012120889292822944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,6144,0.015811555915408664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,4096,0.013712888790501488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,6144,0.023580445183648005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,5120,0.011125333607196808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,5120,0.013959111438857185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,3584,0.011728888584507836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,51200,0.11183467176225449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,5120,0.021586666504542034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,4096,0.009919111099508073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,3072,0.010355555348926121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,4096,0.012455999851226807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,4096,0.019268444842762418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,3584,0.008956444760163626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,2560,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,3584,0.011752888560295105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,3072,0.008421333299742805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,3584,0.018241778016090393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,3072,0.010933332973056369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,2048,0.007689778175618913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,2560,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,2560,0.00981777740849389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,1536,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,2560,0.015864888827006023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,2048,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,2048,0.008448888858159384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,2048,0.014767110347747803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,7168,0.01310755560795466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,1024,0.005084444251325396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,1536,0.006909333169460297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,768,0.00424177779091729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,1024,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,1536,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,1024,0.007272889216740926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,512,0.0032088889016045462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,1024,0.011241777903503843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,768,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,768,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,256,0.0028693332440323303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,768,0.011069333387745751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,512,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,128,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,256,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,256,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,256,0.00903200027015474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,64,0.0021564445147911706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,128,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,3072,0.017289777596791584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,7168,32,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,128,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,128,0.007697777615653143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,65536,0.13129511144426134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,64,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,32,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,7168,1536,0.007373332977294922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,51200,0.10190666384167141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,65536,0.06659822331534491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,16384,0.04013155566321479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,51200,0.0873875551753574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,51200,0.05385155479113261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,512,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,16384,0.020900444851981267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,65536,0.13900800546010336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,16384,0.03256977929009332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,512,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,12288,0.03227555420663621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,51200,0.11159200138515896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,10240,0.02849777870708042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,16384,0.044928001032935254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,12288,0.01739377776781718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,8192,0.021292444732454088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,12288,0.02492444382773505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,10240,0.015462223026487561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,10240,0.021669333179791767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,12288,0.03653600149684482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,7168,0.02019733356104957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,8192,0.013793778088357715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,8192,0.019102222389645047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,10240,0.03217333224084642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,6144,0.01768977774514092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,7168,0.012716444830099741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,8192,0.027797334724002417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,7168,0.016938666502634685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,65536,0.11564889219072129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,7168,0.025610667135980394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,6144,0.011774222056070963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,6144,0.015498666299713982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,6144,0.023193778263198003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,5120,0.013747555514176687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,5120,0.021088000800874498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,3584,0.011335999601417117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,4096,0.012145777543385824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,4096,0.009166222479608323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,4096,0.018983110785484314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,3584,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,3072,0.01017066670788659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,3584,0.011351111034552256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,3584,0.018007111218240526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,2560,0.008631111019187504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,3072,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,3072,0.010655110908879174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,3072,0.016809778081046212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,2048,0.00778399987353219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,2560,0.009633777870072259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,2560,0.007624000310897827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,5120,0.014132445057233175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,2560,0.015415110521846347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,1536,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,2048,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,2048,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,1024,0.005193777796294955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,5120,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,2048,0.014659555421935188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,1536,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,1536,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,768,0.0042168887125121225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,1536,0.0124177775449223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,1024,0.006558222075303395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,512,0.0032106666929192017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,1024,0.011380444798204633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,768,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,768,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,256,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,768,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,512,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,512,0.010415111151006486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,256,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,256,0.005435555759403441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,256,0.009024000002278222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,64,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,128,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,32,0.0021946666141351065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,128,0.005320888840489917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,4096,0.012834666503800286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,64,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,65536,0.11353422535790338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,32,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,51200,0.09086133374108209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,6144,1024,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,65536,0.1091537806722853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,65536,0.06386844317118327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,16384,0.03481866584883796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,51200,0.0516844458050198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,512,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,51200,0.08484533098008896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,12288,0.028897778855429754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,65536,0.13860710461934408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,6144,128,0.0024577778660588795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,51200,0.11082044574949478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,10240,0.02530133393075731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,16384,0.04403733213742574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,12288,0.016910221841600206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,12288,0.023927999867333308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,8192,0.018044445249769423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,12288,0.03597866826587253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,10240,0.015015999476114908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,10240,0.020894222789340552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,7168,0.015457777513398064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,10240,0.03106133474244012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,8192,0.013328889177905189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,8192,0.018544889158672757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,6144,0.013462222284740873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,8192,0.027220444546805486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,7168,0.012457778056462606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,7168,0.016570667425791424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,5120,0.01197866681549284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,6144,0.011335111326641507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,6144,0.015079110860824585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,7168,0.02494488822089301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,4096,0.010397333237859938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,6144,0.02277244461907281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,16384,0.031027555465698242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,5120,0.010254222485754225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,3584,0.010939555863539377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,5120,0.013343111508422427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,16384,0.02070577773782942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,4096,0.008420444197124904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,4096,0.011996444728639392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,3072,0.009911999934249455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,4096,0.018917333748605516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,3584,0.011014222270912595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,128,0.007585778004593319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,2560,0.008402666283978356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,3072,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,3072,0.009927999642160203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,3072,0.016576889488432143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,2048,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,2560,0.008649778034951951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,2560,0.014429334137174817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,2048,0.008294222255547842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,1536,0.006254222244024277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,2048,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,2048,0.01368977791733212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,1536,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,1024,0.004560889055331548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,1536,0.0070168889231152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,5120,0.021017778250906203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,1536,0.011758222348160215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,3584,0.008303999900817871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,768,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,1024,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,3584,0.017758222089873422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,1024,0.0069066666894488865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,512,0.0034479999707804788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,1024,0.011049778097205691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,768,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,768,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,256,0.0028524444335036804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,768,0.010335110955768162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,512,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,512,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,256,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,512,0.009694221946928237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,256,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,64,0.002151111140847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,32,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,256,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,128,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,128,0.0053297777970631914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,65536,0.0888764460881551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,128,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,64,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,5120,32,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,51200,0.07356622483995226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,65536,0.06891466511620416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,16384,0.029743111795849268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,51200,0.04923911227120293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,51200,0.053011556466420494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,65536,0.13815377818213567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,12288,0.023856888214747112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,16384,0.020290666156344943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,51200,0.1103102233674791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,16384,0.04373866650793287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,12288,0.016134222348531086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,2560,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,12288,0.0162791113058726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,12288,0.03497066762712266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,10240,0.014456000592973499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,8192,0.015104888214005364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,10240,0.014588443769348992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,10240,0.030677331818474665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,7168,0.01344622257683012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,8192,0.012827555338541666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,8192,0.013073777986897362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,65536,0.06154133213890923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,8192,0.026710222164789837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,6144,0.01185955521133211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,7168,0.011733333269755045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,7168,0.01180266671710544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,5120,0.010309333602587381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,6144,0.010398222340477837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,6144,0.011001778145631155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,16384,0.0196142229768965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,6144,0.022873777482244704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,4096,0.009282666775915358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,5120,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,5120,0.010080888867378235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,5120,128,0.00245066670080026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,10240,0.020785777105225455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,4096,0.008631111019187504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,4096,0.009280000295903947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,3072,0.007480888730949826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,4096,0.01851111153761546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,3584,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,3584,0.008250666989220513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,2560,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,3584,0.01755911111831665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,3072,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,3072,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,2048,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,3072,0.016539555456903245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,2560,0.007592000067234039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,2560,0.0148062225845125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,1536,0.004583111239804162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,2048,0.006606222026877933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,7168,0.024453333682484094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,2048,0.013336000343163809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,1536,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,1024,0.00349866681628757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,1536,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,1536,0.012375999655988483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,5120,0.020621332857343886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,768,0.0031786666562159858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,1024,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,1024,0.006325333482689328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,1024,0.011342222491900126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,512,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,768,0.006112888869312074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,768,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,768,0.010399111443095738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,256,0.0024560000747442245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,512,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,512,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,2560,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,512,0.010031999813185798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,128,0.0021911110315057966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,256,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,256,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,64,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,256,0.00925244473748737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,32,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,128,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,128,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,65536,0.08067288663652208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,64,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,4096,3584,0.0081404447555542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,4096,32,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,51200,0.06559022267659505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,65536,0.06604622470008002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,65536,0.0603075557284885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,16384,0.02715733316209581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,51200,0.05069244570202298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,51200,0.048600001467598804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,65536,0.13829599486456975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,12288,0.021439111895031396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,16384,0.019891555110613506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,16384,0.01936177743805779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,16384,0.04292000002331204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,2048,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,12288,0.016175111134847004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,10240,0.017682666579882305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,12288,0.015827554795477126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,8192,0.017255999975734286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,128,0.005872000008821487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,12288,0.03417244553565979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,10240,0.014295111099878946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,7168,0.012796444197495779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,10240,0.014096889230940076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,8192,0.012364444633324942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,8192,0.012770666844315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,8192,0.0266942216290368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,7168,0.011139555937714048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,7168,0.011696889168686338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,5120,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,7168,0.024453333682484094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,6144,0.010864888628323873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,6144,0.010016889207892949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,4096,0.008832000195980072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,6144,0.02281066609753503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,5120,0.009304888546466827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,5120,0.009678222239017487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,3584,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,5120,0.020282665888468426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,51200,0.11012888616985744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,4096,0.008367111285527548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,4096,0.008358221915033128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,3072,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,4096,0.018858666221300762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,3584,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,3584,0.007813333637184566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,2560,0.006227555374304454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,3072,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,3584,0.01647733317481147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,3072,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,2048,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,3072,0.015480889214409722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,2560,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,2560,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,1536,0.004541333350870344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,10240,0.030682666434182063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,2560,0.013779555757840475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,2048,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,1024,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,2048,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,2048,0.012786666552225748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,1536,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,768,0.0031555555760860443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,1536,0.006071110980378256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,1024,0.006301333506902059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,1536,0.012238221863905588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,1024,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,1024,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,768,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,256,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,768,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,768,0.01034755590889189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,512,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,512,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,128,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,512,0.010405333505736457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,256,0.0058808889653947614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,256,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,64,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,256,0.009169778062237633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,32,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,128,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,128,0.004975111120276981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,65536,0.0705253349410163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,128,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,64,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3584,32,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,6144,0.011357333097192975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,51200,0.056904892126719155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,65536,0.06292000081804064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,65536,0.059207108285692006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,16384,0.020920000142521326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,51200,0.04734577735265096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,51200,0.04905955659018623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3584,512,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,65536,0.13764888710445827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,12288,0.017120000388887193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,16384,0.01924088928434584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,16384,0.019135110908084445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,51200,0.11005689038170709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,10240,0.014752000570297241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,16384,0.043262223402659096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,12288,0.015820443630218506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,8192,0.0136835558546914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,12288,0.015236443943447538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,10240,0.014122666584120857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,12288,0.03410844339264764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,10240,0.013787555197874704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,10240,0.03007022208637661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,8192,0.011767110890812345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,8192,0.012463111016485425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,6144,0.010614222122563256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,8192,0.026752000053723652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,7168,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,5120,0.009675555759006077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,7168,0.024326221810446844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,6144,0.010262221925788457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,6144,0.010476444330480365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,4096,0.00847822262181176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,6144,0.022629333866967097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,5120,0.008990222381220924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,5120,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,3584,0.007415999968846639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,5120,0.020099555452664692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,4096,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,4096,0.008345777789751688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,3072,0.00675733346078131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,4096,0.017218665944205392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,3584,0.008020444048775567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,3584,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,2560,0.006044444524579578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,3584,0.01685422162214915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,3072,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,3072,0.007231999602582719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,2048,0.0052968888647026485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,2560,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,7168,0.012101333174440594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,2560,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,1536,0.004218666503826777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,2560,0.013552889227867126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,2048,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,1024,0.0032213332338465583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,2048,0.01278133359220293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,7168,0.011019555230935415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,1536,0.006705777926577463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,1536,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,768,0.0028906667398081887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,1536,0.012140444583363004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,1024,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,512,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,1024,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,768,0.006169777777459886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,1024,0.011062222222487131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,768,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,256,0.0025031111306614345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,768,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,512,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,512,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,128,0.0021582223061058256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,512,0.010063110954231685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,256,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,256,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,64,0.0018853334089120228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,3072,32,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,256,0.009295110901196798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,128,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,128,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,128,0.007781333393520779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,3072,0.015511110424995422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,64,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,32,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,51200,0.049265778727001615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,65536,0.06009511152903239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,3072,2048,0.006568000134494569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,65536,0.05826666620042589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,16384,0.021067554752031963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,51200,0.047264890538321606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,65536,0.13760356108347574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,12288,0.016591111818949383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,51200,0.10996088716718887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,16384,0.018868444694413077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,16384,0.04195377892918057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,10240,0.014514666464593677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,12288,0.015279110934999255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,12288,0.014861333701345654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,12288,0.033956444925732084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,10240,0.013178666432698568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,8192,0.012614222036467658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,10240,0.013342222405804528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,10240,0.030181331766976252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,65536,0.05927911069658068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,7168,0.011710222396585675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,8192,0.011071110765139261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,8192,0.026202667090627883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,6144,0.010512000156773461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,7168,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,7168,0.011055111057228513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,7168,0.023982221881548565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,5120,0.009188444250159794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,51200,0.04601866669125027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,6144,0.009697777529557547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,6144,0.010371555884679159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,4096,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,16384,0.018602665927675035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,6144,0.021990221407678392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,5120,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,5120,0.009517333573765224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,5120,0.019606222709019978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,4096,0.00847822262181176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,4096,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,3072,0.0068417779273457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,4096,0.017489777670966256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,3584,0.0079644446571668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,3072,0.007719111111429002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,3584,0.01647200021478865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,2560,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,3072,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,8192,0.012039110892348819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,3072,0.0151573336786694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,2048,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,2560,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,2560,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,1536,0.004192000048028098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,2560,0.013457777599493662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,2048,0.006994666324721442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,2048,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,2048,0.013067555924256643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,1024,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,1536,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,768,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,1536,0.006246222390068903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,1024,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,3584,0.0074551113777690465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,512,0.002564444517095884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,1024,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,1024,0.011056000159846412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,768,0.006224888894293044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,256,0.002401777853568395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,3584,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,512,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,768,0.01074222226937612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,512,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,128,0.0021520000365045336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,512,0.009703999592198266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,256,0.005639111002286275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,256,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,64,0.001993777851263682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2560,32,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,128,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,128,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,65536,0.04961155520545112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,128,0.007954667011896769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,64,0.005196444276306364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,32,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,51200,0.040822221173180476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,65536,0.056477334764268666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,65536,0.05766489108403524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,16384,0.017684444785118103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,1536,0.011752888560295105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,51200,0.04498933421240913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2560,768,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,12288,0.013965333501497904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,16384,0.01862044466866387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,16384,0.041963554090923734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,12288,0.014548444085650973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,12288,0.014544000228246054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,10240,0.014192889134089151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,12288,0.03405155407057868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,256,0.009329777624871995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,8192,0.012236444486512078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,10240,0.013172444370057849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,10240,0.013117333253224691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,10240,0.029785778787400987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,8192,0.011704000333944956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,7168,0.010671111444632212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,8192,0.011048888994587792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,8192,0.02650044361750285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,65536,0.1376106606589423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,6144,0.009879111415810054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,7168,0.010433778166770935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,51200,0.11042400201161702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,7168,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,7168,0.023923556009928387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,16384,0.01809955636660258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,5120,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,6144,0.009353777600659264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,6144,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,4096,0.010066666536860997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,5120,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,6144,0.02176622218555874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,5120,0.009007111191749573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,3584,0.009191999832789103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,5120,0.019313777486483257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,4096,0.008290666672918532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,4096,0.007925333248244392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,3072,0.00776444458299213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,4096,0.016999999682108562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,3584,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,3584,0.0075288886825243635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,2560,0.007090667055712805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,3584,0.016138666205936007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,3072,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,3072,0.00720266666677263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,2048,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,3072,0.0151573336786694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,2560,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,2560,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,2560,0.013478221992651621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,51200,0.04563733273082309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,2048,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,2048,0.006523555351628198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,1024,0.0032106666929192017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,2048,0.012789333032237159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,1536,0.006791111081838608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,1536,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,768,0.0030177777840031516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,1024,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,1536,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,1024,0.005632888939645555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,512,0.002564444517095884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,1024,0.011376000112957425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,768,0.005538666827811136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,768,0.006127111199829314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,256,0.0024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,768,0.010478222535716163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,512,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,128,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,512,0.009756444229020013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,256,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,256,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,64,0.002151111140847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,256,0.009451555709044138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,128,0.00564444437623024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,128,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,65536,0.0383857786655426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,128,0.007358222372002072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,1536,0.003896000070704354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,64,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,32,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,65536,0.055276443560918175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,65536,0.05580800109439426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,16384,0.01587199999226464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,65536,0.1377075513203939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,51200,0.04456622070736355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,51200,0.04371910956170824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,12288,0.013461333182122974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,51200,0.10938044389088948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,16384,0.01756000022093455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,16384,0.017681777477264404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,10240,0.011701333026091257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,2048,512,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,12288,0.01346933344999949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,12288,0.014397333065668741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,8192,0.011000889043013254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,12288,0.03362044360902574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,10240,0.012096000214417776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,2048,32,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,10240,0.012729778057999082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,7168,0.010159999959998662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,10240,0.03005689051416185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,8192,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,8192,0.010718222293588849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,6144,0.009470221896966299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,7168,0.010323555933104621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,8192,0.025977777110205755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,51200,0.030826667944590252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,5120,0.00868800034125646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,7168,0.009953777823183272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,7168,0.023023999399609033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,6144,0.009338666995366415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,4096,0.007850666840871176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,6144,0.02109688851568434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,5120,0.008991999758614434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,5120,0.008637333081828224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,3584,0.006526222245560751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,5120,0.01864266726705763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,4096,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,4096,0.007873777714040544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,3072,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,4096,0.016857778032620747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,3584,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,3584,0.007445333732499017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,16384,0.04166133205095927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,2560,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,3072,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,3072,0.007120889094140794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,2048,0.004585777719815572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,3072,0.014816888504558139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,2560,0.007328000333574083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,2560,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,1536,0.003866666720973121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,2560,0.013763555222087435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,2048,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,2048,0.0063999998900625445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,1024,0.0032151111712058387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,2048,0.012709333664841123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,1536,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,1536,0.00599911105301645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,6144,0.00963199966483646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,768,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,1536,0.011711111499203576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,1024,0.0064177778032090925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,1024,0.00565155554148886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,512,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,768,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,1024,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,256,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,768,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,768,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,512,0.005728000154097875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,128,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,512,0.005296000176005893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,512,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,256,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,256,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,64,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,3584,0.015840000576443143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1536,32,0.0018808889306253856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,256,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,65536,0.027894222074084814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,128,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,128,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,51200,0.023014222582181294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,64,0.004942222187916438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1536,32,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,65536,0.0547599991162618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,65536,0.05270933442645603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,16384,0.012672888735930124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,51200,0.04417333337995741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,51200,0.04213688770929972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,12288,0.011065777805116443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,16384,0.017320000463061862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,51200,0.10900177558263142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,16384,0.016240888171725802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,16384,0.041783110962973706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,10240,0.010085333552625444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,12288,0.01350844403107961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,12288,0.013106666505336761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,8192,0.011751111182901593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,12288,0.03330311179161072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,10240,0.012192000117566852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,7168,0.011071110765139261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,10240,0.02896266513400608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,8192,0.010951999988820819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,8192,0.010513777534166971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,6144,0.009490666290124258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,8192,0.025003555748197768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,7168,0.01033866653839747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,7168,0.00998222248421775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,5120,0.008959111240175035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,7168,0.022450667288568284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,6144,0.009645333720578088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,6144,0.009082666701740688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,4096,0.00794488853878445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,6144,0.02095111045572493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,65536,0.13702932993570963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,5120,0.008652444514963362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,3584,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,5120,0.018580444984965853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,4096,0.0083155557513237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,4096,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,3072,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,4096,0.017224889662530687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,3584,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,3584,0.015836444165971544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,2560,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,3072,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,3072,0.00700444479783376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,3072,0.01481599940194024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,128,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,2048,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,2560,0.007268444531493717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,2560,0.006672888994216919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,2560,0.013752889302041797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,1536,0.0047564444442590075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,2048,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,2048,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,2048,0.01310755560795466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,1024,0.003903111235962974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,5120,0.008977777428097194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,1536,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,1536,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,768,0.0032195556494924757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,1536,0.012102222277058495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,1024,0.006061333335108227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,1024,0.005650666852792104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,512,0.002842666581273079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,768,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,1024,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,768,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,256,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,768,0.010452444354693094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,512,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,512,0.005422222117582957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,128,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,512,0.009654222263230218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,256,0.004991111242108875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,256,0.009659555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,64,0.0021520000365045336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,128,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,1024,32,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,128,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,65536,0.027298667364650305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,1024,32,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,65536,0.051404443052079946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,51200,0.022775999373859827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,65536,0.054321779145134814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,3584,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,65536,0.1375839975145128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,16384,0.01055555542310079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,10240,0.012049777640236748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,51200,0.043756445248921715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,51200,0.041434665520985924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,12288,0.009307555854320526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,256,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,51200,0.1092640029059516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,16384,0.01616533266173469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,10240,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,16384,0.016321778297424316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,128,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,16384,0.041048887703153826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,12288,0.013408888545301227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,8192,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,12288,0.013083555632167392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,10240,0.012096000214417776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,12288,0.03287377622392442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,10240,0.011751111182901593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,8192,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,10240,0.028770665327707928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,6144,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,8192,0.010382222632567087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,7168,0.010016889207892949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,8192,0.02510133385658264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,5120,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,7168,0.009773333039548662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,6144,0.009316444396972656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,4096,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,6144,0.009199110998047723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,5120,0.008985777695973715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,6144,0.020285333196322124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,3584,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,5120,0.008401778009202745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,5120,0.018518222702874076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,4096,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,4096,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,3072,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,4096,0.016579555140601266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,3584,0.007826666865083907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,3584,0.0074551113777690465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,2560,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,3584,0.01586666703224182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,3072,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,3072,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,3072,0.01513155632548862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,2560,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,2560,0.006697777658700943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,1536,0.0038311113086011675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,2560,0.013433777623706393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,2048,0.0070222218831380205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,2048,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,1024,0.00313688897424274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,2048,0.013064888616402945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,1536,0.00665155549844106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,1536,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,768,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,1536,0.011720889144473605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,7168,0.022660444180170696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,1024,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,512,0.002492444382773505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,768,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,1024,0.011013333168294696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,768,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,768,0.010059555371602377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,512,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,256,0.0021679999513758556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,512,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,512,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,256,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,128,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,2048,0.004335111214054955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,256,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,64,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,256,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,128,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,32,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,65536,0.021335111724005804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,128,0.004973333328962326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,128,0.00758755538198683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,64,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,32,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,65536,0.053880890210469566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,65536,0.05087466703520881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,768,1024,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,65536,0.13688445091247559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,51200,0.03875199953715006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,51200,0.04109866751564874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,51200,0.1090480022960239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,16384,0.015847999188635085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,12288,0.008974221845467886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,16384,0.015944888194402058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,16384,0.04075110952059428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,10240,0.008533333738644918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,12288,0.01348444405529234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,12288,0.0131217779384719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,12288,0.03257955445183648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,768,7168,0.008316444853941599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,8192,0.008731555607583787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,10240,0.012074666718641916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,10240,0.011736888852384357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,7168,0.007334222396214803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,10240,0.028788444068696763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,8192,0.0107351111041175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,51200,0.0169493340783649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,6144,0.006302222195598815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,7168,0.010067555639478896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,7168,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,16384,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,7168,0.022827555735905964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,5120,0.006899555524190267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,6144,0.009389333426952362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,6144,0.008978666530715095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,4096,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,6144,0.020283554991086323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,5120,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,5120,0.008375110725561777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,3584,0.006304889089531369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,5120,0.018548443913459778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,4096,0.00831200016869439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,4096,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,3072,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,4096,0.016511110795868766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,3584,0.007823111282454597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,3584,0.007330666813585494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,2560,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,3584,0.01578044394652049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,3072,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,2048,0.004232888834344016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,3072,0.014535110857751636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,2560,0.007113777928882175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,8192,0.010330666270520952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,2560,0.006666666517655055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,1536,0.0038542221817705366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,8192,0.025006221400366888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,2048,0.006899555524190267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,2048,0.0063573333124319715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,1536,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,2048,0.012444444828563266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,1536,0.005984888722499211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,768,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,1536,0.011528000235557556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,1024,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,1024,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,512,0.0025253333151340485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,1024,0.011057777537239922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,768,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,768,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,256,0.002208000048995018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,768,0.010051555931568146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,512,0.0057884446448749965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,512,0.005285333428117964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,128,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,512,0.00998755544424057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,3072,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,256,0.004967111266321606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,64,0.0018195555441909367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,32,0.0018844445132546956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,256,0.008979555633332994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,65536,0.015009777413474189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,128,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,2560,0.013270222478442721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,128,0.0075804442167282104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,64,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,512,32,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,51200,0.013162666724787818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,512,1024,0.0031537777847713898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,65536,0.0495395561059316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,16384,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,51200,0.03791466686460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,65536,0.13647200001610651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,51200,0.03937155670589871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,12288,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,51200,0.10880621936586167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,16384,0.015847111741701763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,16384,0.04007377889421251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,12288,0.013410666750537025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,12288,0.012785777449607849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,10240,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,12288,0.03250755535231696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,10240,0.01202755586968528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,8192,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,10240,0.011710222396585675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,128,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,10240,0.028190222051408555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,7168,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,8192,0.01036977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,8192,0.024999999337726172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,6144,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,7168,0.00998577806684706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,65536,0.046988444195853345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,7168,0.022282666630215112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,5120,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,6144,0.009337777892748514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,6144,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,16384,0.015631111131774057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,4096,0.00600355573826366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,5120,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,5120,0.018287110659811232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,4096,0.007375111182530721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,4096,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,3584,0.005852444304360284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,4096,0.01644000079896715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,3584,0.007722666694058313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,3072,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,3584,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,3584,0.015798222687509324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,2560,0.0052879999081293745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,7168,0.009520000053776635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,3072,0.007608888877762689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,3072,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,2048,0.00489688871635331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,3072,0.014484445254007975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,2560,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,2560,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,1536,0.0035617777870761026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,6144,0.020273778173658583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,2560,0.013452444639470843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,2048,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,5120,0.008929777476522658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,1024,0.002871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,2048,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,1536,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,2048,0.012439111040698158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,768,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,1536,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,1024,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,512,0.00249422217408816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,1536,0.011395555403497485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,1024,0.005640888793600931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,768,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,1024,0.011023110813564725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,256,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,768,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,512,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,768,0.010248888697889116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,128,0.0018746666610240936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,512,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,8192,0.01070577816830741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,256,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,512,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,64,0.001867555495765474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,256,0.005160888863934411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,256,32,0.0017804443422291016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,65536,0.011370666325092316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,256,0.009010666774378883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,128,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,51200,0.01051999959680769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,128,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,64,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,65536,0.04622755448023478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,65536,0.04712355468008253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,51200,0.03736089004410638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,65536,0.1357084380255805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,51200,0.10760888788435195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,16384,0.01552177800072564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,16384,0.015954666667514376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,16384,0.04007911019855075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,12288,0.012729778057999082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,12288,0.013094222380055321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,10240,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,12288,0.03219466739230686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,10240,0.012073777616024017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,128,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,10240,0.011403555671374003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,8192,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,10240,0.02812088860405816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,256,32,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,8192,0.010675555302037133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,8192,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,8192,0.02437244521247016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,51200,0.03803199860784743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,6144,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,7168,0.00963288876745436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,12288,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,7168,0.00998577806684706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,5120,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,7168,0.022309333086013794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,6144,0.008884444832801819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,6144,0.009496000077989366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,4096,0.005285333428117964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,6144,0.02022133270899455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,5120,0.008672888908121321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,5120,0.008276444342401292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,3584,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,5120,0.018255111243989732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,4096,0.008040000167157915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,4096,0.007682667010360294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,3072,0.004852444347408083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,4096,0.016556445095274184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,3584,0.007819555699825287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,3584,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,2560,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,3584,0.01553600033124288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,3072,0.0074142225914531285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,16384,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,2048,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,3072,0.014475555883513557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,2560,0.00720977783203125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,1536,0.003539555602603488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,2560,0.01315822203954061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,2048,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,2048,0.006813333266311222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,1024,0.002847111059559716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,2048,0.01278133359220293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,1536,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,1536,0.011439111497667102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,1024,0.006031111296680238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,1024,0.005533333453867171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,768,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,1024,0.011010666688283285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,512,0.0026782221264309357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,768,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,768,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,768,0.010019555687904358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,256,0.0021199999998013177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,512,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,512,0.005690666536490123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,7168,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,128,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,256,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,256,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,2560,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,64,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,256,0.009184888667530483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,128,32,0.0017911110901170308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,128,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,128,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,65536,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,51200,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,128,0.007099555598364935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,64,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,1536,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,12288,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,32,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,10240,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,65536,0.04725777771737841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,51200,0.03769777880774604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,16384,0.015450666348139444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,12288,0.012726222475369772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,10240,0.011729777687125735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,6144,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,8192,0.009988444546858469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,5120,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,7168,0.009480888644854227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,512,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,4096,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,3584,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,6144,0.008909333083364699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,3072,0.004569777597983678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,5120,0.008047999607192146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,2560,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,4096,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,2048,0.004615111069546806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,3584,0.007246221933099959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,1536,0.0035004446076022256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,16384,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,2560,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,1024,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,768,0.002567111204067866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,2048,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,1536,0.0058044443527857465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,512,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,8192,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,256,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,1024,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,7168,0.005285333428117964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,768,0.0052800000541739995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,64,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,512,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,32,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,256,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,65536,0.01073688848151101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,128,0.004601777841647466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,128,3072,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,64,0.004560889055331548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,51200,0.008970666262838576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,16384,0.005655111124118169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,12288,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,32,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,10240,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,65536,0.047247111797332764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,8192,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,16384,0.015455999308162265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,51200,0.038309332397249006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,7168,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,64,3072,0.007012444237867991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,6144,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,10240,0.01143022212717268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,8192,0.01035288886891471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,7168,0.009700444837411245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,4096,0.004541333350870344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,6144,0.008956444760163626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,3584,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,5120,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,64,128,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,2560,0.004527111020353106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,3584,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,2048,0.00480266660451889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,3072,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,1536,0.0034977777136696708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,2560,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,1024,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,2048,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,768,0.002528888897763358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,1536,0.005798222290145026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,512,0.0021911110315057966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,1024,0.005375111268626318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,256,0.0019502221710152095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,12288,0.012479999827014076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,128,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,512,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,64,0.001777777862217691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,256,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,32,0.0017884444031450485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,128,0.00462311092350218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,64,0.004820444517665439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,4096,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,32,0.004514666481150521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,16384,0.3777182102203369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,16384,0.20093244976467559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,51200,0.6038666831122504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,96,32,768,0.005535111245181825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,12288,0.2796400123172336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,51200,1.1490177578396268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,3072,0.004635555462704765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,16384,0.13566133711073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,12288,0.14868444866604275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,10240,0.20684711138407388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,12288,0.1001360019048055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,8192,0.17254666487375894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,10240,0.12579200002882215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,7168,0.1491911146375868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,10240,0.09242755836910671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,51200,0.4172826608022054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,8192,0.0705582234594557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,7168,0.09337244431177776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,7168,0.07383377684487237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,8192,0.10652799738778009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,5120,0.108751999007331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,6144,0.08142577939563327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,6144,0.0539991127120124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,96,32,5120,0.004592000196377437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,5120,0.04694044589996338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,3584,0.0784319970342848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,5120,0.06671910815768771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,3072,0.06799822383456759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,4096,0.05454044540723165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,3584,0.03950400153795878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,3584,0.04901777704556783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,2560,0.05727644761403402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,3072,0.04222844375504387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,2048,0.04783822099367777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,3072,0.03135022189882066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,1536,0.03643022312058343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,2560,0.027406222290462915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,2560,0.036991112762027316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,2048,0.02637600070900387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,2048,0.03192355566554599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,1536,0.01885688967174954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,768,0.020289777053727042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,1536,0.03971111112170749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,1024,0.015322667029168872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,6144,0.12816266218821207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,512,0.013783999615245394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,1024,0.030115554730097454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,768,0.013725333743625216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,256,0.008375999828179678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,768,0.023354666100607976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,4096,0.09205155240164863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,128,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,256,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,512,0.01665866706106398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,256,0.01239999963177575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,64,0.005111999809741974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,128,0.011158222125636207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,32,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,128,0.011381333072980246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,64,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,65536,32,0.01513066722287072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,65536,1024,0.027196443743175928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,512,0.012540444731712341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,65536,0.6277599864535862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,65536,4096,0.039275556802749634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,51200,0.8555742369757758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,16384,0.2774622175428602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,12288,0.20907644430796304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,16384,0.15964621967739528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,16384,0.11231111155615912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,12288,0.11832000149620904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,65536,0.41959378454420304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,10240,0.17602488729688856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,10240,0.10358044173982407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,10240,0.07078755564159818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,51200,0.33565510643853086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,8192,0.08535466591517131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,7168,0.12695733706156412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,65536,1.093384000990126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,8192,0.0583475563261244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,6144,0.10792711046006943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,7168,0.04976888828807407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,6144,0.04478933413823446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,5120,0.09170755412843491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,7168,0.07231110996670194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,5120,0.03819200065400865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,4096,0.07470844189325969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,6144,0.06314222017923991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,5120,0.05342400074005127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,4096,0.0443173348903656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,3584,0.06626755661434598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,4096,0.032645334800084434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,3072,0.057950225141313344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,3584,0.029042667812771265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,8192,0.14326488971710205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,3584,0.0395671096112993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,2048,0.03988177908791436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,3072,0.026570666167471144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,3072,0.03487022386656867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,12288,0.08332355817159016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,1536,0.03138844503296746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,2560,0.022472888231277466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,2560,0.03030222323205736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,2048,0.022643556197484333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,1024,0.023345778385798138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,768,0.01686844395266639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,1536,0.016081778539551627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,2048,0.026329777306980554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,512,0.011839110818174152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,1536,0.033399111694759794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,1024,0.0135057775510682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,256,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,768,0.012330667012267642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,51200,0.47567378150092227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,1024,0.024624889095624287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,512,0.011019555230935415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,128,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,768,0.018375999397701688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,64,0.004307555655638377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,32,0.004597333156400257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,512,0.014187556174066333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,128,0.00996977753109402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,256,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,64,0.010055111514197456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,51200,256,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,128,0.010009778042634329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,51200,2560,0.04873244298828972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,65536,0.33461954858567977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,65536,0.22493688265482584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,51200,0.25959822866651744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,16384,0.084824886586931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,12288,0.06659644179873996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,65536,0.14489955372280544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,51200,0.17209155029720732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,10240,0.05452355411317614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,16384,0.04140977727042304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,51200,0.11374666955735947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,16384,0.0602088901731703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,8192,0.04736444354057312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,12288,0.03272711237271627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,7168,0.03978133201599121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,10240,0.031137777699364558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,6144,0.03433510992262099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,8192,0.03451822201410929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,7168,0.021721776988771226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,7168,0.029517332712809246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,5120,0.02972888946533203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,6144,0.026496888862715826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,6144,0.019297776950730216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,4096,0.025028443998760645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,51200,32,0.013096888860066732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,5120,0.017419555120997958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,3584,0.02162044412559933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,5120,0.02290311124589708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,4096,0.015249777171346875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,4096,0.019162666466501024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,3072,0.019188443819681805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,3584,0.01425244410832723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,2560,0.015827554795477126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,3584,0.01719199948840671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,3072,0.014890667464998035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,2048,0.012924444344308643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,3072,0.015495111544926962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,2560,0.012392000191741519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,1536,0.010384889112578498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,12288,0.04553777641720242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,10240,0.039339555634392634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,2048,0.012179555164443122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,2560,0.013838222457302941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,8192,0.023992000354660883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,2048,0.012466666599114736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,768,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,1536,0.009538667069541084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,512,0.005139555368158552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,1024,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,1536,0.013998222019937305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,768,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,256,0.003198222153716617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,1024,0.01146755533085929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,768,0.009743111001120673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,512,0.00702311098575592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,256,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,512,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,64,0.002809777855873108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,32,0.0028560000161329904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,256,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,16384,128,0.006816000160243776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,128,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,64,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,16384,32,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,65536,0.2739857832590739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,65536,0.17302933004167345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,51200,0.21743467119004992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,16384,0.07093333535724215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,65536,0.12730133533477783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,51200,0.13205777274237737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,51200,0.10149066978030735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,16384,0.03727644350793626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,10240,0.0449795557392968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,16384,0.04774844315316942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,12288,0.03618844350179037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,128,0.0028560000161329904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,12288,0.029374221960703533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,8192,0.038219554556740656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,7168,0.03294577863481309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,10240,0.025965332984924316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,6144,0.028777778148651123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,10240,0.03138666682773166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,8192,0.02195288903183407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,7168,0.01980533368057675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,5120,0.024718221690919664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,7168,0.023944000403086346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,4096,0.02150933278931512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,6144,0.018106665876176622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,5120,0.01625333395269182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,6144,0.02128000060717265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,3584,0.018231999542978074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,5120,0.0181022220187717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,4096,0.014480888843536377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,3072,0.015793777174419828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,4096,0.015499555402331881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,2560,0.013295999831623502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,3584,0.01351111133893331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,3584,0.014224888549910652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,12288,0.05461600091722277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,2048,0.011511999699804517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,3072,0.013171555267439948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,2560,0.011651555697123209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,1536,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,2560,0.012056888805495368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,2048,0.010775110787815519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,1024,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,1536,0.008638222184446123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,2048,0.01074222226937612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,768,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,1536,0.012339555554919772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,1024,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,512,0.004685333205593957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,1024,0.009413333402739631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,768,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,768,0.008212444682916006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,512,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,128,0.0029324444217814338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,512,0.006760889043410619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,256,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,256,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,64,0.0025235555238193935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,128,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,32,0.0025439999169773525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,128,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,16384,1024,0.008055999875068665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,64,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,65536,0.21344978279537627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,32,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,51200,0.16719467110104033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,65536,0.16353866789076063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,65536,0.12601155704922146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,51200,0.1267768939336141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,16384,0.059953775670793324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,12288,256,0.0034977777136696708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,12288,0.04668000009324816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,12288,8192,0.027679110566775005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,16384,0.03534844517707825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,10240,0.037815110550986394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,51200,0.09822755389743382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,16384,0.044885334041383534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,8192,0.032059556908077665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,12288,0.027874665127860174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,12288,0.034613334470325045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,10240,0.025948445002237957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,10240,0.03035199973318312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,8192,0.02120977805720435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,6144,0.023956444528367784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,7168,0.019357333580652874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,12288,3072,0.012655110822783576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,5120,0.020847999387317236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,7168,0.022622222701708477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,6144,0.017475555340449016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,4096,0.018056000272432964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,6144,0.020064888728989493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,5120,0.015836444165971544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,3584,0.015403555499182807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,4096,0.013920888304710388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,5120,0.017280888226297166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,3072,0.013863110707865821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,4096,0.015156444576051501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,3584,0.013099555340078143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,2560,0.011954666839705573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,3584,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,3072,0.012799111505349478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,2048,0.010285333626800114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,2560,0.010779555473062726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,1536,0.008460444708665213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,2560,0.011701333026091257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,2048,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,1024,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,2048,0.010681777364677854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,1536,0.008358221915033128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,7168,0.02755822241306305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,1536,0.011062222222487131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,1024,0.007924444145626491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,8192,0.026032888227038916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,1024,0.008736000292830998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,256,0.0031973332580592898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,768,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,768,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,512,0.006646222124497096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,128,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,256,0.006303999986913469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,64,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,256,0.005976888868543837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,10240,128,0.0064186664919058485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,32,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,128,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,64,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,65536,0.17115378379821777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,32,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,3072,0.012764444781674279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,51200,0.13280888398488364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,65536,0.12310132715437148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,16384,0.046647999021742076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,65536,0.07094755437639025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,768,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,12288,0.036791112687852644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,51200,0.05693066782421536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,16384,0.022633777724372015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,10240,512,0.004057777838574516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,10240,0.03031022349993388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,16384,0.03422133458985223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,8192,0.02606755495071411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,10240,0.016171556380059984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,12288,0.026939556002616882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,7168,0.022340445054901972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,8192,0.01460533340771993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,10240,512,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,8192,0.01993333299954732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,6144,0.019426667028003268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,5120,0.016511110795868766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,7168,0.014111111561457316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,6144,0.0124177775449223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,4096,0.014288889037238227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,6144,0.01605955594115787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,5120,0.011448889142937131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,3584,0.012411555482281579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,5120,0.014074666632546319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,4096,0.010670222342014313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,4096,0.012608889076444836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,3072,0.010952889091438718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,3584,0.010055999788973067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,3584,0.011610666910807291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,2560,0.00963199966483646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,51200,0.09348711040284897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,3072,0.009300444689061906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,2048,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,3072,0.011023999916182624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,2560,0.008179555336634317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,2560,0.010047111246320937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,12288,0.018570666511853535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,2048,0.008993777963850234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,10240,0.02334844403796726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,1024,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,1536,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,768,0.0042453333735466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,1024,0.006736889067623351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,7168,0.017636444833543565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,512,0.0032115555885765287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,1024,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,768,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,256,0.0028044444819291434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,512,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,512,0.005958222266700532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,128,0.0025351111673646504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,256,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,256,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,64,0.002465777720014254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,128,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,32,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,128,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,8192,1536,0.006907555378145642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,64,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,8192,2048,0.007887110941939885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,65536,0.1627475553088718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,32,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,51200,0.12913155555725098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,1536,0.008945778012275696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,16384,0.04288355509440104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,65536,0.06787821981641981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,51200,0.09205600288179185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,65536,0.11786222457885742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,51200,0.05519288778305054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,12288,0.03347200155258179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,10240,0.02810933192571004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,8192,768,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,16384,0.022329777479171753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,12288,0.018012444178263348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,8192,0.02369244396686554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,7168,0.02071377800570594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,12288,0.025835555460717943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,10240,0.015760888655980427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,10240,0.02234222160445319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,6144,0.017921777235137094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,8192,0.014107555150985718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,5120,0.015501333607567681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,7168,0.01315199997689989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,8192,0.020008888509538438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,6144,0.012095111111799875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,7168,0.01721155477894677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,6144,0.015876443849669564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,5120,0.011056889262464313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,4096,0.009696000152164036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,5120,0.01387555566098955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,3072,0.0103493332862854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,4096,0.01274222218328052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,2560,0.008985777695973715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,3584,0.01166044423977534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,3072,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,2048,0.00790666706032223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,3072,0.010776888993051318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,2560,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,1536,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,2560,0.009755555954244401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,2048,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,1024,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,2048,0.008350222474998897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,16384,0.03331022130118476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,768,0.0041804446114434134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,1536,0.008793777889675563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,1024,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,512,0.0032826666202810076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,1024,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,768,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,256,0.002581333327624533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,768,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,4096,0.013351110948456658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,512,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,128,0.00254488881263468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,3584,0.01167377746767468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,512,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,64,0.002426666725012991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,256,0.00564533347884814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,7168,32,0.002463999928699599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,256,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,128,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,3584,0.008800889054934183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,128,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,65536,0.14002754953172472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,64,0.005791999813583162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,7168,32,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,51200,0.11130932966868083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,65536,0.06566311253441705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,16384,0.04049333267741733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,65536,0.11185510953267415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,12288,0.03171644608179728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,51200,0.0530008872350057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,51200,0.08625244432025486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,16384,0.02131822208563487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,10240,0.027804444233576458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,8192,0.021741333934995864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,12288,0.01738933391041226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,12288,0.024505777491463557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,10240,0.015418666932317944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,7168,0.019444444113307528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,10240,0.021477333373493616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,8192,0.013770666387346057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,6144,0.017516444126764934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,8192,0.018970666660202872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,7168,0.012700444294346703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,5120,0.014727999766667684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,7168,0.01682933337158627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,6144,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,4096,0.012431110772821637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,5120,0.010710222025712332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,3584,0.010753778119881948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,4096,0.009030222064918941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,4096,0.012080888781282636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,3584,0.008626666333940294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,3072,0.010326222413116032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,3584,0.011330666641394297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,2560,0.008606221940782335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,3072,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,3072,0.010424888796276517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,2048,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,7168,1536,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,2560,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,16384,0.032280001375410296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,1536,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,2048,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,1024,0.004618666652176115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,2048,0.0079644446571668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,1536,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,768,0.00423200014564726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,1536,0.008519111408127679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,1024,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,512,0.0031431110368834603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,1024,0.007070221834712558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,6144,0.015488889482286243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,768,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,256,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,5120,0.013432000246312885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,128,0.0024053334361977046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,512,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,256,0.0058711109062035876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,32,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,256,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,128,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,128,0.005283555636803309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,65536,0.1104924413892958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,64,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,32,0.006167111297448476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,6144,2560,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,51200,0.08705688847435845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,16384,0.03290133343802558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,65536,0.10732355382707383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,12288,0.027748444014125403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,51200,0.05097688900099861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,51200,0.0839795536465115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,16384,0.020403555697864957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,10240,0.023935111032591924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,16384,0.0307626658015781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,12288,0.016791999340057373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,8192,0.021379555265108745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,10240,0.015005333556069268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,10240,0.02091911103990343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,6144,768,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,8192,0.01330044451687071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,6144,0.013544888959990608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,6144,64,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,7168,0.012423999607563019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,7168,0.01641244524055057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,6144,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,6144,0.015013333823945789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,4096,0.010673777924643623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,5120,0.013222222526868185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,5120,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,3584,0.009343999955389235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,65536,0.06320622232225206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,4096,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,3072,0.009639110830095079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,4096,0.011755555868148804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,3584,0.008027555214034187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,2560,0.008601778083377415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,3584,0.01110755569405026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,3072,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,12288,0.023680888944202002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,2560,0.007438222567240397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,2048,0.007350222104125553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,7168,0.01849600010448032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,2560,0.008458666503429413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,1536,0.0061919999619325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,2048,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,8192,0.018568888306617737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,1024,0.004531555705600315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,1536,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,2048,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,768,0.003914666672547658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,5120,0.011807999677128263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,1024,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,512,0.0035253332720862497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,1024,0.006903999795516332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,256,0.002814222127199173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,512,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,512,0.005721777677536011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,256,0.00565244464410676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,128,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,256,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,64,0.0021137777302000257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,128,0.005816000203291575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,5120,32,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,128,0.005314666777849197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,3072,0.00980088859796524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,64,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,65536,0.08681066830952962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,32,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,51200,0.07100711266199748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,65536,0.06110577450858223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,65536,0.06702578067779541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,16384,0.029118223322762385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,51200,0.04895022180345324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,1536,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,51200,0.051911999781926475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,12288,0.02312177750799391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,5120,768,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,10240,0.02016444504261017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,16384,0.01959466603067186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,5120,768,0.006340444501903322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,16384,0.020271999968422782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,12288,0.016174222032229107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,7168,0.013390222357379066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,10240,0.014525334040323893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,12288,0.016176000237464905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,6144,0.011750222080283694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,8192,0.012825777961148156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,10240,0.014476444986131458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,5120,0.01036444471942054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,7168,0.011728000309732227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,8192,0.013045333325862885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,4096,0.00944800012641483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,6144,0.01053155544731352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,7168,0.011737777955002256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,3584,0.008271111382378472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,5120,0.00941777808798684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,6144,0.011071999867757162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,3072,0.0074444446298811175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,4096,0.008561777571837107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,5120,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,2560,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,3584,0.008143111235565608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,3584,0.008301333420806462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,2048,0.005688888745175467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,3072,0.007343110938866933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,2560,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,1536,0.004286222159862518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,2560,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,2048,0.007061333292060428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,1024,0.0035075553589397003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,1536,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,1024,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,1536,0.00701155596309238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,8192,0.015322667029168872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,1024,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,512,0.002809777855873108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,768,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,256,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,512,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,768,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,128,0.0022257777551809945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,512,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,256,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,64,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,256,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,32,0.0022231110682090125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,128,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,4096,0.008947555389669206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,65536,0.07680355840259127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,32,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,51200,0.06227199898825752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,65536,0.06464178032345243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,65536,0.060164444976382785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,16384,0.027279999521043565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,2048,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,4096,768,0.0031822222388452957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,12288,0.021391999390390184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,51200,0.05022844341066149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,16384,0.019368888603316415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,16384,0.01943555474281311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,10240,0.018569777409235638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,12288,0.01609866652223799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,12288,0.01570933394961887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,8192,0.016578666037983365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,10240,0.014382221632533602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,128,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,7168,0.01256622208489312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,10240,0.014126222994592456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,8192,0.012410666379663678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,8192,0.012734221915404001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,4096,3072,0.007874666816658443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,6144,0.011227555572986603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,7168,0.011356444822417365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,5120,0.009943111075295342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,7168,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,6144,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,4096,0.009160888691743216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,5120,0.00903911143541336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,3584,0.007761778102980719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,5120,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,4096,0.00833777752187517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,3072,0.007116444408893585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,4096,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,3584,0.008064889245563084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,51200,0.047897779279285006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,2560,0.0063475556671619415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,3584,0.007620444728268518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,3072,0.007771555748250749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,2048,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,3072,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,2560,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,1536,0.004239999999602635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,2560,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,2048,0.007051555646790399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,1024,0.003526222374704149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,2048,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,768,0.0031555555760860443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,1536,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,1024,0.006317333214812809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,512,0.0028604444944196274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,1024,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,4096,64,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,256,0.0024862223201327855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,768,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,512,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,512,0.005290666802061929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,128,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,256,0.005682666682534748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,64,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,256,0.005191111316283544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3584,32,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,128,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,128,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,65536,0.06885600090026855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,64,0.005095110999213325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,32,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,51200,0.05413866705364651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,65536,0.05912355581919352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,65536,0.06193866994645861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,16384,0.022190221481853064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,12288,0.018044445249769423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,1536,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,51200,0.0466177761554718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,51200,0.04831555485725403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,10240,0.01555999947918786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,16384,0.018703111343913607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3584,768,0.006167111297448476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,10240,0.013927110367351107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,12288,0.01513777838812934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,7168,0.012172444827026792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,10240,0.013629333012633853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,8192,0.0120666664507654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,6144,0.010842666857772403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3584,6144,0.010728889041476779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,7168,0.010703999963071613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,8192,0.012319111161761813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,5120,0.009657777845859528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,7168,0.011403555671374003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,6144,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,5120,0.00906666699382994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,6144,0.01038933296998342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,3584,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,5120,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,4096,0.008395555946562026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,3072,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,4096,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,3584,0.008168000313970778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,2560,0.006221333311663733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,3584,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,3072,0.007683555285135905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,2048,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,3072,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,16384,0.01905777719285753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,2560,0.007323555648326874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,1536,0.004268444246715969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,2560,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,12288,0.01584533353646596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,2048,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,1024,0.0033173333439562055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,8192,0.013422222601042854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,1536,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,2048,0.0063475556671619415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,1024,0.006412444429265128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,1536,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,512,0.0025902222841978073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,1024,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,768,0.006100444330109491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,256,0.0021946666141351065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,512,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,768,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,128,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,4096,0.008594666918118795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,512,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,256,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,128,0.00573600000805325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,128,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,65536,0.06344444221920438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,64,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,3072,32,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,51200,0.05174933208359612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,65536,0.05744355254703098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,16384,0.020213334096802604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,65536,0.059317330519358315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,51200,0.045999109745025635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,51200,0.046520888805389404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,12288,0.016181333197487723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,16384,0.018908444378111098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,768,0.0028764444092909494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,10240,0.014096889230940076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,16384,0.018204443984561496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,12288,0.015281778242852954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,8192,0.012225777738624148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,12288,0.014757333530320061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,10240,0.013021333350075616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,7168,0.011705777711338468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,8192,0.011184000306659274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,10240,0.013450666434235044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,6144,0.01053244454993142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,8192,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,3072,256,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,7168,0.010498666928874122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,5120,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,4096,0.008216000265545314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,3072,32,0.0020639999873108333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,7168,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,6144,0.00963288876745436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,3584,0.00739822205570009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,5120,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,6144,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,4096,0.00834222220712238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,3072,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,5120,0.009119999905427298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,2560,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,3584,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,4096,0.00777511133088006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,2048,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,3584,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,3072,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,1536,0.003949333396222857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,2560,0.007324444750944774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,1024,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,3072,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,2560,0.006698666761318843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,768,0.002875555513633622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,1536,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,2048,0.006545777950021956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,1024,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,1536,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,256,0.002352888799375958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,768,0.005966222120655908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,512,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,768,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,128,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,256,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,512,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,64,0.001820444439848264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,32,0.001960888918903139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,256,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,128,0.005636444522274866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,65536,0.05264977614084879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,64,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,32,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,51200,0.04249155521392822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,65536,0.056087997224595815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,16384,0.016540444559521146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,51200,0.04488177763091194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,51200,0.04452355702718099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2560,2048,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,12288,0.013390222357379066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,16384,0.017606221967273288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,16384,0.018346667289733887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2560,512,0.002543111021320025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,12288,0.014954666296641031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,8192,0.011857777833938599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,1024,0.005720888988839255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,12288,0.014475555883513557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,10240,0.01238933371173011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,7168,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,10240,0.013078221844302284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,8192,0.011056889262464313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,6144,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,8192,0.011711999773979187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,7168,0.010232888989978367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,7168,0.010448888772063784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,5120,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2560,128,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,6144,0.009703999592198266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,4096,0.007751999629868402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,6144,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,3584,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,4096,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,65536,0.0565333366394043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,3072,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,4096,0.007962666451931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,3584,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,2560,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,3584,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,3072,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,2048,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,3072,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,2560,0.007345777418878343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,1536,0.0039022221333450745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,10240,0.012710221939616732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,2560,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,1024,0.0032035555276605818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,2048,0.007015110717879401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,2048,0.006311999840868845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,768,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,1024,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,1024,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,512,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,768,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,768,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,256,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,5120,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,512,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,128,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,5120,0.008341333104504479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,256,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,64,0.0021280000607172647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,256,0.005197333378924264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,2048,32,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,128,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,65536,0.040361778603659734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,128,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,64,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,51200,0.032349334822760686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,32,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,65536,0.05505599909358554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,16384,0.014030221435758801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,65536,0.054515557156668774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,12288,0.013073777986897362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,51200,0.04450311263402303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,2048,1536,0.00675644435816341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,51200,0.042797333664364286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,10240,0.012059555285506778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,2048,1536,0.005976000179847081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,16384,0.017196445001496207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,12288,0.014292443792025248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,8192,0.011032889286677042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,7168,0.010235555469989777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,10240,0.012351999680201212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,12288,0.01421333352724711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,8192,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,6144,0.009317333499590555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,10240,0.012796444197495779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,5120,0.008787555827034844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,7168,0.010335110955768162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,7168,0.009919111099508073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,6144,0.009370666411187913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,4096,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,5120,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,6144,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,3584,0.006424888968467712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,3072,0.005872000008821487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,5120,0.008469333251317343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,4096,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,2560,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,3584,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,4096,0.007895111209816402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,3072,0.007669332954618666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,3584,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,2560,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,1536,0.0038897775941424896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,2048,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,2048,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,16384,0.016768000192112394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,1536,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,1536,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,768,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,1024,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,1024,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,512,0.002532444480392668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,768,0.006070222291681502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,8192,0.010455110834704505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,768,0.005433777968088786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,512,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,512,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,256,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,256,0.005058666484223472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,2048,0.004465777840879229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,128,0.0019075555933846368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,64,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,3072,0.007246221933099959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,32,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1536,128,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,2560,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,65536,0.029320889049106177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,128,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,51200,0.023971555961502924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,1024,0.0031848889258172777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,32,0.005192888693677055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,65536,0.05453777644369337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,16384,0.011382222175598145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,65536,0.05166133244832357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,51200,0.04392710990375943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,51200,0.041101333167817854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,16384,0.01696977847152286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1536,256,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,10240,0.010039110978444418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,12288,0.013376000026861826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,12288,0.013437333206335703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,8192,0.011843555503421359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,10240,0.012054222325483957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,7168,0.011940444509188334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,10240,0.012042666474978128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,8192,0.01068800025516086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,6144,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,8192,0.010484444598356882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,7168,0.010110222631030613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,5120,0.008917333351241218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,7168,0.009831111464235518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,6144,0.00941777808798684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,4096,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,6144,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,5120,0.00885333369175593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,5120,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,4096,0.007693332930405934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,3072,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,3584,0.007322666545708974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,12288,0.009667555491129557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,2560,0.005127999931573868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,16384,0.01613066593805949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,3072,0.007366221812036302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,2048,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,3072,0.007056000332037608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,2560,0.007246221933099959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,1536,0.0046862223082118565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,2560,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,2048,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,1024,0.0035377778112888336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,2048,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,1536,0.006310222049554189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,768,0.0031946665710873077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,1536,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,1024,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,512,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,1024,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,768,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,3584,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,256,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,512,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,4096,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,128,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,256,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,64,0.0021182222084866627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,3584,0.007702222300900354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,128,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,256,0.005205333232879639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,1024,32,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,65536,0.024326221810446844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,128,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,51200,0.019893333315849304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,64,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,32,0.005091555416584015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,65536,0.05430755681461758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,16384,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,65536,0.050759110185835093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,12288,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,51200,0.043474667602115206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,51200,0.04087288843260871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,16384,0.01590666671593984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,10240,0.008651555412345463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,16384,0.01608444419172075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,12288,0.013239111337396832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,8192,0.009122666385438707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,12288,0.012944888737466602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,10240,0.012028444144460889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,7168,0.008263111114501953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,1024,768,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,10240,0.011776000261306763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,8192,0.010729778144094678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,6144,0.007191999918884701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,7168,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,5120,0.007931555310885111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,8192,0.01036977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,7168,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,6144,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1024,512,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,4096,0.0069066666894488865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,5120,0.008970666262838576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,3584,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,6144,0.009121778110663096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,3072,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,5120,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,3584,0.007669332954618666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,2560,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,3584,0.007376000285148621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,3072,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,2048,0.004551110996140374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,3072,0.007130666739410824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,2560,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,2560,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,1536,0.0038382220599386427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,2048,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,1024,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,2048,0.006335999816656113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,1536,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,768,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,1536,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,1024,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,512,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,1024,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,768,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,256,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,768,0.005528000079923206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,512,0.0058088890380329555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,1536,64,0.005063111169470681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,512,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,128,0.001820444439848264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,256,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,64,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,768,32,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,256,0.00519822206762102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,128,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,65536,0.02091377807988061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,4096,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,128,0.005047111047638787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,51200,0.017422222428851657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,64,0.004964444372389051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,16384,0.011298666397730509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,65536,0.05353422297371758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,65536,0.04991288979848226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,51200,0.04037066631846958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,12288,0.009511999785900116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,16384,0.01591111057334476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,51200,0.04038044479158189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,10240,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,16384,0.01549600064754486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,12288,0.013269333375824822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,8192,0.008264888491895463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,12288,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,7168,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,8192,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,10240,0.011771555576059552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,6144,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,8192,0.01032266683048672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,7168,0.010084444450007545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,5120,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,7168,0.009639110830095079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,6144,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,4096,0.00675733346078131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,6144,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,5120,0.008625778059164682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,768,4096,0.008088000118732452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,3584,0.006214222146405115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,5120,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,3072,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,4096,0.008303999900817871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,4096,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,3584,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,768,32,0.004953777624501122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,3072,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,3584,0.007253333098358578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,3072,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,2560,0.007069333559936947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,1536,0.0038853333228164246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,2560,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,2048,0.0069066666894488865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,1024,0.003165333428316646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,1536,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,768,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,1536,0.005979555762476391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,10240,0.011997333003415002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,1024,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,512,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,1024,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,256,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,768,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,512,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,128,0.0019288888821999233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,512,0.0052897776994440295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,256,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,64,0.001828444500764211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,256,0.004956444518433677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,512,128,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,32,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,128,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,65536,0.01835466590192583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,64,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,2560,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,32,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,51200,0.015557333827018738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,512,2048,0.004291555533806483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,16384,0.011066666907734342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,65536,0.04662044511901008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,65536,0.04912177721659342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,12288,0.009308444129096137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,51200,0.03771911064783732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,16384,0.015800888339678448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,51200,0.03758577836884393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,2048,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,10240,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,16384,0.01550044450494978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,12288,0.013418667018413544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,12288,0.012793777717484368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,7168,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,8192,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,512,768,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,8192,0.010285333626800114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,6144,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,7168,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,5120,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,7168,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,6144,0.009369778136412302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,4096,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,6144,0.009005332986513773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,5120,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,3584,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,5120,0.008292444050312042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,4096,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,3072,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,4096,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,3584,0.007844444778230455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,2560,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,3584,0.007248000138335758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,3072,0.007348444726732042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,2048,0.0042160000238153665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,3072,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,8192,0.007330666813585494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,1536,0.003572444534964032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,2560,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,10240,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,1024,0.00315022220214208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,2048,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,10240,0.011419555379284753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,1536,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,768,0.0026586666289303037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,1536,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,512,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,768,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,1024,0.00536088893810908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,256,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,768,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,512,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,512,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,256,0.005789333333571752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,64,0.0018035556293196147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,256,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,128,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,32,0.001828444500764211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,128,0.004838222016890843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,65536,0.023429332507981196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,64,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,65536,0.04640266630384657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,256,32,0.00452622233165635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,2560,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,65536,0.04625511169433594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,2048,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,51200,0.03734755516052246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,12288,0.0069857777820693115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,16384,0.015841777125994366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,256,1024,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,10240,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,12288,0.013169777890046438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,12288,0.012801777985360889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,256,128,0.0018817778262827133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,8192,0.005650666852792104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,10240,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,7168,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,10240,0.011392888923486074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,8192,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,6144,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,7168,0.010039110978444418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,8192,0.010210666391584609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,5120,0.006455111006895701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,7168,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,6144,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,51200,0.020572443803151447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,4096,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,6144,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,3584,0.005642666584915585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,16384,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,5120,0.008272888759771982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,3072,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,4096,0.008085333638721043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,51200,0.037335998482174344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,3584,0.007663999994595845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,2560,0.0047280001971456735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,3584,0.007087111473083496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,3072,0.007357333269384172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,16384,0.015186667442321777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,2048,0.004171555654870139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,3072,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,1536,0.003539555602603488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,2560,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,2048,0.006787555499209299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,1024,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,2048,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,1536,0.006607999818192587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,768,0.0028008888992998335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,1536,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,1024,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,512,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,1024,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,768,0.006063111126422882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,256,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,768,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,512,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,128,0.0019617778145604664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,512,0.004978666702906291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,256,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,64,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,5120,0.00885511106914944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,256,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,128,32,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,81,128,128,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,65536,0.010658666491508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,128,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,51200,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,16384,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,64,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,12288,0.0053022222386466126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,32,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,10240,0.005532444351249271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,65536,0.045956446064843066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,8192,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,51200,0.0369742214679718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,2560,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,16384,0.01550222271018558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,6144,0.005647999958859549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,12288,0.012415111064910889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,10240,0.011505777637163797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,8192,0.010071111222108206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,4096,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,7168,0.009321778184837764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,3584,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,5120,0.00831822223133511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,3072,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,4096,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,2560,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,3584,0.007096000015735626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,2048,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,3072,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,1536,0.003539555602603488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,1024,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,128,4096,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,2560,0.0063662222690052455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,768,0.002549333290921317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,512,0.00249422217408816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,1536,0.005966222120655908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,256,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,768,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,128,0.0018560000591807896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,512,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,64,0.001791999985774358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,5120,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,256,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,65536,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,128,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,6144,0.009027555584907532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,51200,0.00942133367061615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,64,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,16384,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,12288,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,65536,0.04609510964817471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,10240,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,51200,0.03733955489264594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,8192,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,16384,0.015273777974976433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,12288,0.012458666331238218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,7168,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,6144,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,10240,0.011252444651391772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,5120,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,2048,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,8192,0.01036977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,4096,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,7168,0.009370666411187913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,3584,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,1024,0.005285333428117964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,3072,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,6144,0.00904088881280687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,2560,0.004865777575307422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,64,7168,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,2048,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,5120,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,1536,0.0035342222286595237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,4096,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,1024,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,3072,0.006908444480763541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,2560,0.006646222124497096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,768,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,512,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,2048,0.006024888820118374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,256,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,1536,0.005642666584915585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,128,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,1024,0.005624889085690181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,64,32,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,768,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,81,32,32,0.0018684443914228016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,512,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,256,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,128,0.004621333132187526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,64,0.004874666531880696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,32,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,16384,0.36804978052775067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,16384,0.2034977806939019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,81,32,3584,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,51200,0.6009839905632867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,51200,1.1353706783718533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,12288,0.276859548356798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,12288,0.14888177977667913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,16384,0.1353671153386434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,10240,0.2150017817815145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,51200,0.3928951157463922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,10240,0.13131199942694768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,12288,0.10092088911268447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,8192,0.1781608925925361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,8192,0.10736533006032307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,10240,0.09199200073877971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,7168,0.15467733807033962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,8192,0.07018133004506429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,7168,0.09151289198133682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,16384,0.28481777509053546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,7168,0.06332355737686157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,8192,0.14777778254614934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,10240,0.18504088454776338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,12288,0.20903200573391387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,6144,0.07869244284100003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,5120,0.11113333702087402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,7168,0.12767022185855442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,6144,0.054711110062069364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,5120,0.0660524434513516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,4096,0.09267289108700222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,5120,0.04583200150065952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,4096,0.055128001504474215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,3584,0.07811822493871053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,6144,0.11283377806345622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,5120,0.09481510851118301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,4096,0.07955733272764418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,3072,0.06816355387369792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,3584,0.048708445496029325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,3072,0.03202222122086419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,3072,0.04244622257020739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,2560,0.058576001061333545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,3584,0.07147110833062066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,3072,0.06366755565007527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,2560,0.036841776635911726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,2048,0.04711022310786777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,2560,0.027160889572567407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,2560,0.05592177973853218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,1536,0.03661600086424086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,4096,0.039502223332722984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,2048,0.0240346673462126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,2048,0.031810667779710554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,2048,0.048935999472936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,1024,0.026538666751649644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,1536,0.020405333903100755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,3584,0.03760177890459696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,6144,0.1330151160558065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,768,0.020579554968410067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,1536,0.04746577805942959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,1024,0.014986667368147107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,512,0.013646221823162503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,768,0.023629332582155865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,1024,0.0373342235883077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,768,0.02901866700914171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,256,0.008341333104504479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,512,0.012322666744391123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,512,0.01688088807794783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,512,0.02494844463136461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,256,0.011035555766688453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,128,0.006515555497672822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,256,0.012474666866991254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,256,0.02055199940999349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,64,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,128,0.011055111057228513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,128,0.011403555671374003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,65536,32,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,128,0.016120000018013846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,1536,0.03954755597644382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,64,0.01149777736928728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,32,0.015505777464972602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,65536,1024,0.028586665789286297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,768,0.013743999931547375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,65536,0.6271760198805068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,51200,0.8741804228888618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,65536,1.0856800079345703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,51200,0.4752391179402669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,51200,0.8532204627990723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,16384,0.27658843994140625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,65536,0.40828267733256024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,16384,0.1592577828301324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,12288,0.20733243889278838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,51200,0.3188409010569255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,16384,0.10711289114422268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,12288,0.11818133460150824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,10240,0.1731111076143053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,12288,0.08289955721961127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,10240,0.10019466612074111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,8192,0.14179555575052896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,16384,0.23150134086608887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,10240,0.07010666529337566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,8192,0.08520533641179402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,7168,0.12501422564188638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,8192,0.056964443789588086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,12288,0.1793973313437568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,7168,0.07214755482143827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,7168,0.05075110991795858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,6144,0.10678666830062866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,10240,0.1478239960140652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,8192,0.11831022633446588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,5120,0.09078400002585517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,6144,0.0631964471605089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,6144,0.04395022326045566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,7168,0.10613777902391221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,4096,0.07388622230953641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,5120,0.05331822236378988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,5120,0.037949333588282265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,6144,0.09291466739442612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,4096,0.032709333631727434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,3584,0.06498399708006117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,5120,0.08075910806655884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,4096,0.044166223870383374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,3072,0.05699200100368924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,4096,0.06857422325346205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,3584,0.039665778477986656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,3584,0.028583112690183852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,2560,0.04775733417934842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,3584,0.06200977828767565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,3072,0.02593244446648492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,3072,0.034670223792394005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,2048,0.039100444979137845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,3072,0.054959999190436475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,2560,0.02353777819209629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,1536,0.030866665972603693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,2560,0.04861155483457777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,2048,0.020636444290479023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,2048,0.02643288837538825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,2048,0.0432604451974233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,1024,0.02235377828280131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,1536,0.01594044433699714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,1536,0.03244444396760728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,51200,0.7087519963582357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,1536,0.03619822197490268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,1024,0.014178666803571912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,1024,0.024384000235133704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,512,0.011424000064531961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,1024,0.03185688787036472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,768,0.012131555212868584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,768,0.025461332665549383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,256,0.007516444557242923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,512,0.010834666589895884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,512,0.014613333675596448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,65536,0.9543013042873807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,768,0.02125777800877889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,128,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,512,0.021589333812395733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,256,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,64,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,256,0.010450666977299584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,51200,128,0.009923555784755284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,32,0.00387999994887246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,256,0.0188400000333786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,2560,0.03031377659903632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,128,0.010739555789364709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,51200,128,0.014053333136770459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,64,0.010122666756312052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,51200,32,0.013192888763215808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,65536,0.32268532117207843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,51200,768,0.016188444362746347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,51200,0.24999822510613334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,65536,0.2262649006313748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,16384,0.08181511031256782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,65536,0.1466426716910468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,51200,0.17197867234547934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,12288,0.06472088893254598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,16384,0.060193777084350586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,16384,0.04127733243836297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,10240,0.052763554784986705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,51200,0.11104710896809895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,12288,0.04576355550024244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,12288,0.032356444332334734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,16384,0.06007733609941271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,12288,0.04638311266899109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,8192,0.0450799994998508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,10240,0.03935911258061727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,10240,0.029137776957617864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,10240,0.04110844598876105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,7168,0.03837955660290188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,8192,0.03413866625891791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,8192,0.024119999673631456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,6144,0.033505777517954506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,7168,0.029490666257010564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,7168,0.021386666430367365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,51200,0.15627200073666042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,7168,0.032219555642869734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,5120,0.02852977646721734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,6144,0.026524444421132404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,6144,0.02090311050415039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,4096,0.024699555502997503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,5120,0.017303999927308824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,6144,0.03000977635383606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,4096,0.019128888845443726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,4096,0.015503111812803479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,5120,0.026084444589085046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,3584,0.020908445119857788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,4096,0.02309333284695943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,3584,0.014118222726715935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,3072,0.018333334061834548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,3584,0.017449777987268236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,3584,0.021300445000330608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,2560,0.015210666590266757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,3072,0.013221333424250284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,3072,0.015533333023389181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,8192,0.03655022382736206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,3072,0.01924088928434584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,2048,0.012431110772821637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,2560,0.012069332930776807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,2560,0.013793778088357715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,65536,0.20073955588870576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,1536,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,2560,0.018276444739765592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,2048,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,2048,0.01242844429281023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,1024,0.007771555748250749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,2048,0.01680444512102339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,1536,0.009775111244784461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,768,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,1536,0.013995556367768182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,1536,0.015503999259736804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,1024,0.011678222152921887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,512,0.0047280001971456735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,1024,0.013991999957296582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,768,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,768,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,256,0.0031679999083280563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,768,0.012028444144460889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,512,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,512,0.008275555239783393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,128,0.0028080000645584534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,512,0.01054044481780794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,256,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,64,0.0024986666523747975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,256,0.010045333041085137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,128,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,128,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,64,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,32,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,65536,0.24933245446946886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,5120,0.022893332772784766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,65536,0.17352622085147432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,51200,0.19709511597951254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,65536,0.1294977797402276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,51200,0.13220000267028809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,16384,256,0.006565333240562015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,128,0.00665511108107037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,16384,0.06996177964740329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,16384,32,0.00254844439526399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,51200,0.10202666785981919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,16384,0.04706933432155185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,16384,0.0387662218676673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,16384,0.05154044429461161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,12288,0.029639111624823675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,10240,0.045200887653562755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,12288,0.036306666003333196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,65536,0.1639288928773668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,12288,0.04126044445567661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,8192,0.03835555579927232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,51200,0.13075289461347792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,10240,0.025648888614442613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,7168,0.03314222229851617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,10240,0.03155644403563605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,10240,0.03663466705216302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,8192,0.027391110857327778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,8192,0.02180355621708764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,6144,0.02865866820017497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,8192,0.03225955698225234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,7168,0.019645333290100098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,7168,0.023972445064120825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,5120,0.02481066683928172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,7168,0.029137776957617864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,6144,0.02127733329931895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,6144,0.018106665876176622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,4096,0.021203555994563635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,12288,0.05474399858050876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,6144,0.02662311163213518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,5120,0.016248888439602323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,3584,0.018180444836616516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,5120,0.023440000083711412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,4096,0.01443733274936676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,4096,0.015829333000712924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,3072,0.01571822166442871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,4096,0.021312000022994146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,1024,0.008207110895050896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,3584,0.014116444521480136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,3584,0.014104889498816596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,2560,0.013111111190583972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,3584,0.019301333361201815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,3072,0.012744888663291931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,2048,0.011008000208271874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,3072,0.013078221844302284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,2560,0.011381333072980246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,3072,0.01846933364868164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,2560,0.016843555702103507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,1536,0.008959111240175035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,2048,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,2048,0.010766222245163388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,1024,0.007353777686754863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,2048,0.016164445214801364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,1536,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,1536,0.012262221839692859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,768,0.006162666612201267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,1536,0.01478666729397244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,1024,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,5120,0.018518222702874076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,1024,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,512,0.004707555390066571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,1024,0.012749333348539142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,768,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,256,0.003367111086845398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,768,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,768,0.011746666497654386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,512,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,512,0.0069146665434042616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,128,0.0028684443483750024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,512,0.010552888943089379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,256,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,256,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,64,0.002542222125662698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,12288,32,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,256,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,128,0.006535999890830781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,128,0.005966222120655908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,2560,0.011957333319716983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,128,0.008207110895050896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,64,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,65536,0.21236711078219941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,12288,32,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,51200,0.1660791105694241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,65536,0.16379200087653265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,16384,0.05821688969930013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,65536,0.11955555280049641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,51200,0.1264151069853041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,12288,0.04721422327889336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,51200,0.09724177916844685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,16384,0.03633155425389608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,16384,0.04497333367665609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,10240,0.03762755460209317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,16384,0.04956266615125868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,12288,0.02803022331661648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,12288,0.034653332498338484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,8192,0.03246311015552945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,12288,0.03980888923009237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,10240,0.024681778417693243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,10240,0.030200888713200886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,51200,0.12328622076246475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,65536,0.15544266170925564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,8192,0.026016889346970454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,8192,0.021702221698231165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,10240,0.03534844517707825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,6144,0.024808888634045918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,8192,0.030972444348865088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,7168,0.01924977699915568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,5120,0.020967110991477966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,7168,0.022778666681713525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,6144,0.018228444788191054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,4096,0.018269333574506972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,7168,0.02776266634464264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,6144,0.025663110944959853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,5120,0.015676443775494892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,3584,0.015469332536061605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,5120,0.017299556069903903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,5120,0.023019555542204116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,4096,0.014311111635631986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,4096,0.015200000670221118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,3072,0.013461333182122974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,4096,0.020078221956888836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,3584,0.01311911145846049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,3584,0.01402933398882548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,2560,0.011866666376590729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,3072,0.01202044470442666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,3072,0.01309511148267322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,3072,0.018039999736679923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,2048,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,2560,0.011693333586057028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,2560,0.010886222124099731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,7168,0.027954667806625366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,1536,0.008634666601816813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,2048,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,2048,0.010671111444632212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,2048,0.015501333607567681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,1024,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,1536,0.010871999793582492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,6144,0.020204444726308186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,1536,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,1536,0.014127999544143677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,1024,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,1024,0.008584000170230865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,1024,0.012094222009181976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,768,0.007469333708286285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,512,0.004377777791685528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,768,0.007088888850477006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,768,0.011382222175598145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,512,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,512,0.0068408888247278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,512,0.010063110954231685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,2560,0.016972444123691983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,128,0.002817777709828483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,256,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,256,0.009663110805882348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,128,0.00655377780397733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,128,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,64,0.0025191110455327565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,768,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,32,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,64,0.006186666587988536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,65536,0.17188355657789442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,10240,32,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,3584,0.019039111004935372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,10240,256,0.0032711111836963226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,65536,0.12134310934278701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,65536,0.07083555724885729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,51200,0.13295643859439427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,16384,0.046671112378438316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,256,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,51200,0.09325066539976333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,51200,0.057032889790005155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,12288,0.03636888994110955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,16384,0.03425333234998915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,16384,0.02310133311483595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,10240,0.03010666701528761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,16384,0.045392000012927584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,128,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,12288,0.018400889303949144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,65536,0.13867644468943277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,8192,0.02572533322705163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,12288,0.026974222726292078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,12288,0.036935110886891685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,51200,0.11096533139546712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,10240,0.016406221522225272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,7168,0.022209778428077698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,10240,0.023398222194777593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,8192,0.014552889598740472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,6144,0.019479110836982727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,10240,0.03291022115283542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,7168,0.013613333304723104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,5120,0.016473778420024447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,8192,0.02881422307756212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,7168,0.017694221602545846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,6144,0.01256711118751102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,7168,0.026199999782774184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,6144,0.01607111096382141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,4096,0.014059555199411182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,5120,0.011368888947698804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,5120,0.01407199932469262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,3584,0.012326222327020435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,4096,0.012708444562223224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,5120,0.02165599995189243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,4096,0.019550222489568923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,3072,0.010880000061459012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,3584,0.011622221933470832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,3584,0.009411555197503831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,2560,0.009526222116417354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,3584,0.01828888886504703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,3072,0.008715555899673039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,3072,0.011019555230935415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,2048,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,3072,0.016955556141005624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,2560,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,2560,0.00998755544424057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,1536,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,2560,0.016249777542220224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,8192,0.019916445016860962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,2048,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,2048,0.008913777768611908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,1024,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,2048,0.015164444843928019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,1536,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,1536,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,768,0.004210666649871402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,1536,0.012816888590653738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,1024,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,1024,0.007884444461928474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,6144,0.024108444650967915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,512,0.0031759999692440033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,4096,0.010715555813577441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,1024,0.011542222566074796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,768,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,768,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,512,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,512,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,128,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,512,0.010312000082598792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,256,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,256,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,64,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,256,0.009316444396972656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,32,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,128,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,128,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,128,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,64,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,65536,0.1515831152598063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,8192,32,0.006067555397748947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,51200,0.11960889233483209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,65536,0.11842400497860378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,65536,0.06806933217578463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,51200,0.09091644154654609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,16384,0.045490665568245776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,8192,256,0.002842666581273079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,51200,0.05515911181767782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,768,0.010420444111029306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,16384,0.021891555852360193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,16384,0.033406221204333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,65536,0.1379351086086697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,10240,0.028135110934575398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,51200,0.11078577571445042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,16384,0.044842667049831815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,12288,0.026135111848513286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,12288,0.017873777283562552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,8192,0.02363288899262746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,10240,0.015804444750150044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,12288,0.03654133280118307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,7168,0.02068000038464864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,10240,0.032487110959159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,8192,0.014085332552591959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,8192,0.028134223487642076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,6144,0.01789955629242791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,7168,0.01310755560795466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,7168,0.017258667283587985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,5120,0.015600888265503777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,7168,0.025946666797002155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,6144,0.012080888781282636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,6144,0.015846222639083862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,4096,0.013550221920013428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,12288,0.033598221010631986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,6144,0.023538667294714186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,5120,0.011108444796668159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,3584,0.011746666497654386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,5120,0.013802666631009845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,4096,0.009979555176364051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,5120,0.02139644490347968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,4096,0.012749333348539142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,3072,0.0107342220014996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,4096,0.019337778290112812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,3584,0.01147377739350001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,2560,0.009148444566461775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,3584,0.01792266633775499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,3072,0.008284444610277811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,10240,0.022245332598686218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,2048,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,3072,0.01735555628935496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,2560,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,2560,0.009922666682137383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,1536,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,2560,0.016065778003798593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,2048,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,2048,0.01536266671286689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,1024,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,1536,0.009039999710188972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,1536,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,1536,0.012910222013791403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,1024,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,768,0.003903111235962974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,1024,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,1024,0.01111288865407308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,512,0.003245333416594399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,768,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,3584,0.008856000171767341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,768,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,256,0.0025804444319672054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,768,0.010812444819344414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,512,0.005962666538026597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,3072,0.010684444672531553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,128,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,8192,0.019484443797005545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,512,0.01035822182893753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,256,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,64,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,128,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,256,0.009722666607962714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,7168,32,0.0022071111533376905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,128,0.007576888634098901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,64,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,65536,0.12673866748809814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,32,0.0061928890645504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,65536,0.11137244436475967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,65536,0.06575377782185872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,51200,0.10017333428064983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,512,0.006059555543793573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,51200,0.08711200290256077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,256,0.00554044461912579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,16384,0.038655999634000994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,65536,0.13735555277930364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,51200,0.05306933323542277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,2048,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,12288,0.031994667318132185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,16384,0.021151999632517498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,16384,0.03199555476506551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,51200,0.10980977614720662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,10240,0.027820444769329492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,16384,0.04455555478731791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,12288,0.02461244497034285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,12288,0.017267554998397827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,12288,0.035764445861180626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,8192,0.02295644415749444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,10240,0.015311110350820752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,7168,0.01956622302532196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,8192,0.013645333548386892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,10240,0.03182844320933024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,8192,0.018904889623324077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,8192,0.027268444498380024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,7168,0.01272266689274046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,7168,0.016900445024172466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,5120,0.014155555102560254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,7168,0.02518044412136078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,6144,0.011788444386588203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,6144,0.015343111422326831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,4096,0.01293511109219657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,6144,0.023060444328520033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,5120,0.013608888619475894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,7168,128,0.005320000151793162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,3584,0.012050666742854647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,5120,0.021153777837753296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,4096,0.012281777958075205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,4096,0.018789332773950364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,3584,0.008623999853928884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,3072,0.010081777969996134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,3584,0.011333333121405708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,3584,0.01763288842307197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,10240,0.021537777450349595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,3072,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,2560,0.008928000099129146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,3072,0.010586666564146677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,6144,0.01644799941115909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,3072,0.016524444023768108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,2560,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,2560,0.00903733323017756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,2560,0.015494222442309061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,1536,0.006566222343179915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,2048,0.007328000333574083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,2048,0.00830488900343577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,1024,0.004724444614516364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,2048,0.013752000199423896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,1536,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,5120,0.010458666417333813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,1536,0.008608000146018134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,768,0.0037848887344201407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,1536,0.012154666913880242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,1024,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,512,0.0031582222630580268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,1024,0.011043555206722684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,768,0.006224888894293044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,768,0.006322666588756774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,256,0.002546666603949335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,768,0.010394666757848527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,512,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,512,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,128,0.0023431111541059283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,512,0.00981866651111179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,256,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,256,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,64,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,2048,0.008261333737108443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,256,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,128,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,128,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,128,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,64,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,6144,32,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,65536,0.10919200049506293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,51200,0.09201955795288086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,65536,0.06324711110856798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,1024,0.0064177778032090925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,4096,0.009099555512269339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,16384,0.032815112007988825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,51200,0.08394755257500543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,65536,0.1370933320787218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,16384,0.03086044390996297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,16384,0.02020711037847731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,51200,0.10947466558880276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,16384,0.044032000833087497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,6144,32,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,12288,0.023560000790490046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,12288,0.016770665844281513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,12288,0.03550666570663452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,65536,0.1174595620897081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,8192,0.021485333641370136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,10240,0.014974221587181091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,10240,0.020767110917303298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,10240,0.03122399912940131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,8192,0.01330844478474723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,7168,0.01864266726705763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,8192,0.018535999788178336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,8192,0.027048889133665297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,7168,0.012425777812798819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,51200,0.0510480006535848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,6144,0.013733333183659447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,7168,0.016336000627941556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,5120,0.01198933356338077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,7168,0.025084444218211707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,6144,0.011166222393512726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,12288,0.027611555324660406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,4096,0.010653333531485664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,6144,0.015016888578732809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,6144,0.02277066641383701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,5120,0.013199110825856527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,10240,0.02381422287887997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,3584,0.009405333134863112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,4096,0.008977777428097194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,4096,0.01203733351495531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,3072,0.009858667022652095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,4096,0.018723555737071566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,3584,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,3584,0.010994666980372535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,2560,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,3584,0.017878222796652053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,3072,0.007693332930405934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,3072,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,2048,0.007528000407748752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,3072,0.016167999969588388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,2560,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,1536,0.0061928890645504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,2560,0.008352888955010308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,2048,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,2048,0.007891555627187094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,1024,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,2048,0.013262222210566202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,1536,0.006763555523422029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,1536,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,768,0.004576888763242298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,1536,0.012004444168673621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,1024,0.006414222220579783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,1024,0.006747555401590135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,5120,0.009861333502663506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,512,0.0035013332962989807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,1024,0.011078221930397881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,768,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,256,0.0028986665937635633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,512,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,768,0.010384000009960597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,512,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,512,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,128,0.0022293333378103045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,256,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,64,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,256,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,128,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,5120,32,0.0023680000255505242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,128,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,128,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,65536,0.08717422352896796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,2560,0.015119999647140503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,64,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,32,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,51200,0.07111022207472059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,65536,0.06703111198213366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,65536,0.06084799766540527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,16384,0.02755644420782725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,51200,0.05164177881346809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,51200,0.04885333445337084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,12288,0.023020444644822016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,65536,0.1368675496843126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,5120,768,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,16384,0.01960355540116628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,51200,0.10918400022718643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,16384,0.043340444564819336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,10240,0.020298666424221463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,12288,0.016223111086421542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,8192,0.015154666370815702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,10240,0.014430221584108142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,12288,0.03424800104565091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,10240,0.01442399952146742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,7168,0.013370667066839008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,10240,0.0306328899330563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,8192,0.012810666528013019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,8192,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,6144,0.01180622229973475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,8192,0.02646666599644555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,7168,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,7168,0.011770666473441653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,7168,0.02458577851454417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,5120,0.010566222170988718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,6144,0.01127022256453832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,6144,0.010704889065689512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,6144,0.022446221775478784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,5120,0.009653333160612319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,4096,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,16384,0.02021511064635383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,5120,0.0207360006040997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,5120,0.020823111136754353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,3584,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,4096,0.008384889198674096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,4096,0.009006222089131674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,3072,0.007363555332024892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,12288,0.015896888242827523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,3584,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,3584,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,2560,0.006782222125265334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,3584,0.017528888252046373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,3072,0.007697777615653143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,3072,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,2048,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,3072,0.016135111451148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,2560,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,2560,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,1536,0.004834666848182678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,2560,0.014484445254007975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,2048,0.007012444237867991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,2048,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,1024,0.0035137778355015647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,2048,0.01294311136007309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,1536,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,1536,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,768,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,1024,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,1536,0.012379555238617791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,1024,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,512,0.002842666581273079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,1024,0.011064888702498542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,768,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,768,0.006000888844331105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,256,0.002500444443689452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,4096,0.01828266680240631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,768,0.010095111197895473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,512,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,128,0.0021973333011070886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,512,0.009663999908500248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,256,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,256,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,64,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,256,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,128,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,128,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,128,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,65536,0.07642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,64,0.0053075556125905775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,32,0.0053004444473319584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,51200,0.061536000834570996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,65536,0.06388977501127455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,5120,0.009992889232105678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,65536,0.05996533234914144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,16384,0.024914667010307312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,51200,0.04977600110901726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,65536,0.1362666686375936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,51200,0.04793688986036512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,16384,0.019367999500698514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,51200,0.10932711097929214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,16384,0.019440000255902607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,4096,512,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,10240,0.01666933298110962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,16384,0.04314488834804959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,12288,0.016136889656384785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,8192,0.01574844453069899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,12288,0.015523556205961438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,10240,0.014381332529915703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,12288,0.03408977720472548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,10240,0.014099554883109199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,4096,32,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,7168,0.013927110367351107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,10240,0.03048000070783827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,8192,0.012378666963842181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,6144,0.01277511070171992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,8192,0.012653333445390066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,7168,0.011320888996124268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,5120,0.009947555760542551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,7168,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,6144,0.010070222119490305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,7168,0.024098666177855596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,6144,0.010785777535703448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,4096,0.008602666358153025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,6144,0.022672888305452134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,5120,0.009739555418491364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,3584,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,5120,0.02027555637889438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,4096,0.008415111237102086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,4096,0.008573333422342936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,3072,0.007242666350470648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,12288,0.019281778070661757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,4096,0.0177839994430542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,3584,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,3584,0.016650666793187458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,2560,0.006228444476922353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,3072,0.007323555648326874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,3072,0.014873777826627096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,2560,0.007401777638329401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,2560,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,2048,0.005295111073387994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,2560,0.01421333352724711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,8192,0.026358221968015034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,1536,0.004524444540341695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,2048,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,2048,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,2048,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,5120,0.009019555317031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,1024,0.0035235554807715942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,1536,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,1536,0.006333333336644703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,768,0.003145777723855443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,1536,0.012054222325483957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,1024,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,1024,0.005981333139869902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,3584,0.008167111211352879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,512,0.002834666727317704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,1024,0.010812444819344414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,768,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,256,0.002472888885272874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,3072,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,768,0.010598222414652506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,512,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,128,0.0021946666141351065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,512,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,256,0.005650666852792104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,512,0.010363555616802638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,256,0.00498488876554701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,32,0.0021359999146726397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,256,0.00903200027015474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,128,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,65536,0.06836177905400594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,128,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,64,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,32,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,51200,0.05458044343524509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,65536,0.06177866458892822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,65536,0.058916442924075656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,16384,0.02065955599149068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,51200,0.04818488823042976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,65536,0.1364133358001709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,51200,0.04678044385380215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,12288,0.017119111286269296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3584,768,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,51200,0.1089057789908515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,16384,0.018704889549149405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,16384,0.019225777851210702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,10240,0.014553777045673795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,12288,0.015792888071801927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,12288,0.015146666102939181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,8192,0.012585777375433179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3584,64,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,12288,0.033687111404207014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,10240,0.013968888256284924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,7168,0.012274666792816587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,128,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,10240,0.030135111676322088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,8192,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,6144,0.010936888555685679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,7168,0.011078221930397881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,8192,0.02663555575741662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,7168,0.011369778050316704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,5120,0.00959644466638565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,7168,0.024359999431504145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,6144,0.009687999884287516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,6144,0.010438222024175854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,4096,0.008614222208658854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,6144,0.022286223040686712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,5120,0.009380444056457942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,5120,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,3584,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,5120,0.02016177773475647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,4096,0.008016000191370646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,4096,0.008409777449236976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,3072,0.006909333169460297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,4096,0.01756444407833947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,3584,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,16384,0.04252800014283922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,3584,0.016183111402723525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,10240,0.013707555830478668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,3072,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,3072,0.015112888481881885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,2560,0.007425777614116669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,2048,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,8192,0.012448888685968188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,2560,0.014120888378885059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,1536,0.004553777890072929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,2048,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,2048,0.013171555267439948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,1536,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,1024,0.0033599999215867785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,1536,0.00665511108107037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,1536,0.011695110963450538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,1024,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,3584,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,1024,0.011066666907734342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,2560,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,1024,0.006303999986913469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,3072,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,512,0.0026382222357723448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,768,0.00609688874748018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,768,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,2560,0.006912888752089606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,768,0.01001422190003925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,512,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,512,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,2048,0.00700711127784517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,128,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,512,0.009711110757456886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,256,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,256,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,64,0.0019493332753578823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,256,0.009300444689061906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,128,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,32,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,128,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,65536,0.05882133377922905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,64,0.005172444300519096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,3072,32,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,768,0.0029253332565228143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,51200,0.04719110992219713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,65536,0.0585191117392646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,65536,0.05744000275929769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,16384,0.018975999620225694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,65536,0.13618222872416177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,51200,0.04575377702713013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,51200,0.046338667472203575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,3072,256,0.002491555487116178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,12288,0.016504888733228047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,51200,0.10883111423916286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,16384,0.018777777751286823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,10240,0.014192000031471252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,16384,0.04162133402294583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,12288,0.0151235560576121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,12288,0.03408977720472548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,128,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,10240,0.013613333304723104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,10240,0.029800001117918227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,8192,0.012228444218635559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,7168,0.01187999960449007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,8192,0.026184888349639043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,7168,0.011177777416176267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,6144,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,7168,0.011006222003036074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,7168,0.024442666106753882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,16384,0.017909333109855652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,5120,0.009308444129096137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,6144,0.01034755590889189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,6144,0.009662222531106737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,12288,0.015152000718646579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,4096,0.00815822184085846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,5120,0.009098666409651438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,5120,0.009036444127559662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,8192,0.012120889292822944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,3584,0.0074577778577804565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,5120,0.020222221811612446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,4096,0.00830488900343577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,4096,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,3072,0.006719999843173557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,4096,0.017318222257826064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,3584,0.007931555310885111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,8192,0.011383111278216044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,3584,0.015828443898095023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,2560,0.00600355573826366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,3072,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,3072,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,2048,0.004970666848950916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,3072,0.01518133282661438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,2560,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,2560,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,1536,0.004524444540341695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,2048,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,2048,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,6144,0.02201333310869005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,1536,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,1024,0.0032266666077905228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,1536,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,1536,0.011704888608720569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,768,0.0029004443850782183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,1024,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,1024,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,1024,0.011012444065676795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,512,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,768,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,768,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,768,0.010580444501505958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,256,0.002471999989615546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,512,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,512,0.009654222263230218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,2560,0.01368799971209632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,128,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,256,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,2048,0.0063164445261160536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,256,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,128,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,128,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,128,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,32,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,64,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,65536,0.047889779011408486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,3584,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,32,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,512,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,51200,0.03923199905289544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,65536,0.05602044529385037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,65536,0.05676977501975166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,65536,0.13647645049624973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,256,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,16384,0.016571554872724745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,51200,0.0451822214656406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,51200,0.04457955559094747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,12288,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,51200,0.10930933554967244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,16384,0.018529777725537617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2560,64,0.0019164443429973389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,16384,0.01775911119249132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,10240,0.013060444758998023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2560,10240,0.013366222381591797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,16384,0.04179644584655762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,12288,0.01458755632241567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,8192,0.012059555285506778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,10240,0.01310755560795466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,12288,0.03371289041307237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,10240,0.013159111142158508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,7168,0.010324444207880232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,8192,0.011573333707120685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,10240,0.029512001408471003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,6144,0.009594666461149851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,8192,0.02640266716480255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,7168,0.01033866653839747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,7168,0.01034044474363327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,7168,0.02404088940885332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,5120,0.008841777841250101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,6144,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,6144,0.009658666948477427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,4096,0.011005333728260465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,6144,0.02201866606871287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,5120,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,5120,0.008636444807052612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,3584,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,5120,0.018997333116001554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,4096,0.008286221987671321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,4096,0.00778311077091429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,3072,0.008037333687146505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,3584,0.007729777859316931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,12288,0.014473777678277759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,3584,0.007340444458855524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,2560,0.007111111448870764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,3072,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,3072,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,3072,0.014780445231331719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,2560,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,2560,0.006709333509206772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,8192,0.011072888970375061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,1536,0.004007110993067424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,2560,0.013447999954223633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,2048,0.006405333264006509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,2048,0.007032000356250339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,1024,0.0032195556494924757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,2048,0.012727110750145383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,1536,0.006806222101052602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,1536,0.006026666611433029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,768,0.003019555575317807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,1536,0.011777777638700275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,1024,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,1024,0.0058133333093590206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,512,0.0026657777941889236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,1024,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,768,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,768,0.010061333576838175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,256,0.00244711111817095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,512,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,3584,0.015840889679061044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,512,0.009659555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,2048,0.004527111020353106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,128,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,256,0.005170666509204441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,256,0.005763555566469829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,64,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,256,0.009639999932712978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,2048,32,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,128,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,128,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,65536,0.03622400098376804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,128,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,64,0.00499199993080563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,2048,32,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,51200,0.030085331863827173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,65536,0.05530044436454773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,65536,0.0539591113726298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,16384,0.01535555554760827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,51200,0.044509334696663745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,51200,0.04277599851290385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,51200,0.10813333590825398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,4096,0.017432888348897297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,512,0.0057795556883017225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,12288,0.013733333183659447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,16384,0.01753155555990007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,16384,0.04134666588571336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,10240,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,12288,0.01423111061255137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,12288,0.03392088744375441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,10240,0.0127697777416971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,8192,0.010959111154079437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,10240,0.029151999288135107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,8192,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,65536,0.13599556022220188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,8192,0.025067556235525344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,8192,0.011019555230935415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,7168,0.01034222212102678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,7168,0.010372444159454769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,6144,0.009587555295891231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,7168,0.010384889112578498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,7168,0.02302933401531643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,12288,0.013447110851605734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,5120,0.008722667064931657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,6144,0.009249777429633671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,6144,0.02090399960676829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,10240,0.012085333466529846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,4096,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,5120,0.00872444444232517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,5120,0.008434666527642144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,5120,0.01902844508488973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,3584,0.006544888847404056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,4096,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,4096,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,768,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,3072,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,3584,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,3584,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,3584,0.015829333000712924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,16384,0.01720533271630605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,2560,0.005312888986534542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,3072,0.007691555553012424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,3072,0.007179555793603261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,3072,0.014484445254007975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,2560,0.007358222372002072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,2560,0.00664444433318244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,1536,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,6144,0.009477333062224919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,2560,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,2048,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,2048,0.006411555740568373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,1024,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,2048,0.012483555409643384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,1536,0.006149333384301927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,768,0.0028604444944196274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,1536,0.011753777662913004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,1024,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,1024,0.01109155515829722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,512,0.0025031111306614345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,768,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,768,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,4096,0.016872000363137987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,256,0.0022133334229389825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,768,0.010069333016872406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,512,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,512,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,128,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,512,0.00999466660949919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,256,0.00562755556570159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,2048,0.004475555486149258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,64,0.0018257778137922287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,256,0.009032888544930352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1536,32,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,128,0.004972444640265571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,65536,0.027360000544124182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,128,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,64,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,1536,0.006729777902364731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,51200,0.02429155508677165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,65536,0.05182400014665392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,65536,0.05477866530418396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,1024,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,16384,0.011578666667143503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,65536,0.13596799638536242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,51200,0.04390666551060147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,16384,0.016758221719000075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,51200,0.10831555393007065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,16384,0.015961777832772996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,12288,0.009746666583749983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,16384,0.041625777880350746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,10240,0.010055111514197456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,12288,0.013423110875818463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,256,0.004986666556861666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,12288,0.013448889056841532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,8192,0.012126222252845764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,128,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,10240,0.012079999678664737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,10240,0.01179377817445331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,7168,0.01130222198035982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,8192,0.01073333372672399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,10240,0.02880177895228068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1536,32,0.004971555537647671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,8192,0.010422222316265106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,6144,0.009396444592210982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,8192,0.024661334024535284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,7168,0.009762666291660732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,7168,0.010001777774757808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,5120,0.009386666946940953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,7168,0.022500443789694045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,6144,0.009166222479608323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,6144,0.020583111378881667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,4096,0.0069057775868309875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,5120,0.008363555702898238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,5120,0.008736000292830998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,5120,0.018541332748201158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,3584,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,4096,0.008366222182909647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,4096,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,4096,0.01682844426896837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,3072,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,3584,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,3584,0.007357333269384172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,3584,0.015848888291252982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,2560,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,3072,0.0074684446056683855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,3072,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,3072,0.01475911173555586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,2048,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,2560,0.007270221908887227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,2560,0.006665777828958299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,2560,0.013785777820481194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,1536,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,51200,0.04120355513360765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,2048,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,2048,0.00646577775478363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,1024,0.0038577777643998465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,2048,0.012538666526476542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,1536,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,1536,0.006129777679840724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,768,0.0032266666077905228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,1536,0.011745777395036487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,1024,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,1024,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,512,0.0028248888750871024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,1024,0.01073688848151101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,768,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,12288,0.032984889215893216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,768,0.005439110928111606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,256,0.002491555487116178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,768,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,512,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,512,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,128,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,512,0.009490666290124258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,256,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,256,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,64,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,1024,32,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,256,0.009372444616423713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,128,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,128,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,65536,0.021000888612535264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,6144,0.00963288876745436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,51200,0.02212711175282796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,64,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,1024,32,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,65536,0.05446755554940966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,16384,0.011018666956159802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,65536,0.05101422137684292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,51200,0.04365422328313192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,51200,0.04076177875200907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,65536,0.13498666551378039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,12288,0.009696000152164036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,16384,0.016210666961140104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,51200,0.10763821999231975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,10240,0.008633777499198914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,16384,0.016358221570650738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,16384,0.04088800152142843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,8192,0.00999111102686988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,12288,0.013103110922707452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,10240,0.012053333222866058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,12288,0.03218577636612786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,10240,0.011751111182901593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,7168,0.009033777647548253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,10240,0.028431998358832464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,8192,0.01074222226937612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,8192,0.010355555348926121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,8192,0.024711110525661047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,7168,0.010002666877375709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,7168,0.009698666632175446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,5120,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,128,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,6144,0.009341333475377824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,6144,0.009030222064918941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,4096,0.006674666785531574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,6144,0.02092355489730835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,5120,0.00869777798652649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,5120,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,3584,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,5120,0.018917333748605516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,4096,0.007845333053006066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,4096,0.016833777228991192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,12288,0.013391999734772576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,3072,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,3584,0.00794400026400884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,3584,0.016196444630622864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,3072,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,3072,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,6144,0.008387555678685507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,3072,0.014610666367742749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,7168,0.022333333889643352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,2560,0.006720888945791457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,2560,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,2048,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,2560,0.013759111364682516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,1536,0.0038693332009845306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,2048,0.006697777658700943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,2048,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,4096,0.008284444610277811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,2048,0.01276533305644989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,1024,0.003177777760558658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,1536,0.006198222024573221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,1536,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,768,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,3584,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,1536,0.011722666521867117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,1024,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,1024,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,512,0.0025315555847353404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,768,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,768,0.005423999908897612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,256,0.0021671110557185277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,768,0.010041777458455827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,2560,0.005019555489222209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,512,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,128,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,512,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,256,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,64,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,256,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,128,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,768,32,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,128,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,256,0.008974221845467886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,65536,0.020911999874644812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,128,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,51200,0.01742577718363868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,64,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,65536,0.053048001395331494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,65536,0.05030666788419088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,16384,0.011680000358157687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,51200,0.04042044281959534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,51200,0.04036533170276218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,65536,0.13502577940622965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,12288,0.009572444690598382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,51200,0.10739644368489583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,16384,0.015882665912310284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,16384,0.015632889337009855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,1024,0.010671999719407825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,10240,0.008618666893906063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,16384,0.040419555372662015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,12288,0.012780444489585029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,8192,0.007535999847782983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,512,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,10240,0.012068444656001197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,12288,0.032219555642869734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,10240,0.011663999822404651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,7168,0.0069857777820693115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,10240,0.02847644355561998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,8192,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,6144,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,8192,0.02404088940885332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,7168,0.010259555445777046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,7168,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,5120,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,7168,0.022327999273935955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,6144,0.009628444082207149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,768,32,0.004943111290534337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,4096,0.006085333310895496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,6144,0.009024000002278222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,5120,0.008954666554927826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,6144,0.020431111256281536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,5120,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,3584,0.005986666513813867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,5120,0.018248000078731112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,4096,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,4096,0.007624000310897827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,3072,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,4096,0.01649777756796943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,3584,0.007687110867765214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,3584,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,2560,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,3584,0.015238222148683337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,12288,0.01346933344999949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,3072,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,2048,0.0047493332790003884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,3072,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,3072,0.014463999205165438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,1536,0.0038417776425679526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,2560,0.006853333363930385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,2048,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,2560,0.01351111133893331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,2048,0.006301333506902059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,1024,0.0031991110493739447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,2048,0.013091555900043912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,1536,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,1536,0.011383111278216044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,768,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,1024,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,1024,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,1024,0.011391110718250275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,512,0.0025191110455327565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,768,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,768,0.005290666802061929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,768,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,256,0.0022195554855797025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,512,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,512,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,512,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,128,0.002151111140847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,256,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,256,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,256,0.009003555609120263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,64,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,128,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,8192,0.010341333018408881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,128,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,512,32,0.0018817778262827133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,65536,0.017948443690935772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,128,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,1536,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,51200,0.01585155559910668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,64,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,512,32,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,65536,0.04724533359209696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,65536,0.04920088913705614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,16384,0.007868444754017724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,51200,0.037532445457246565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,51200,0.03867377837498983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,12288,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,51200,0.1069306664996677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,16384,0.015209777487648858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,16384,0.015829333000712924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,10240,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,16384,0.04012622104750739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,12288,0.013053333593739403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,8192,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,10240,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,12288,0.03229599859979417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,10240,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,2560,0.007252444823582967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,7168,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,10240,0.027797334724002417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,8192,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,8192,0.01036177741156684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,6144,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,7168,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,8192,0.024704888463020325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,7168,0.00965244405799442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,5120,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,7168,0.021956443786621094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,6144,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,6144,0.009257777697510189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,4096,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,6144,0.0209840006298489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,5120,0.008683555656009251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,3584,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,65536,0.1342853307723999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,4096,0.008086221913496653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,4096,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,3072,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,4096,0.016549333930015564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,3584,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,3584,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,2560,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,12288,0.013235555754767524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,3584,0.01552088889810774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,3072,0.0074933336840735535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,3072,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,2048,0.0041982221106688184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,3072,0.014493332968817817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,2560,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,2560,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,1536,0.0035439998739295532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,2560,0.013410666750537025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,2048,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,1024,0.003044444446762403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,2048,0.01312711089849472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,1536,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,1536,0.006588444527652528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,768,0.0028435554769304064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,1536,0.011624000138706632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,1024,0.006165333506133821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,1024,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,5120,0.008074666890833113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,1024,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,768,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,768,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,256,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,768,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,512,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,512,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,512,0.010094222095277574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,256,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,128,0.001867555495765474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,256,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,256,0.008836444053384993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,64,0.0018044445249769424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,128,0.005292444593376583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,32,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,128,0.004827555682924059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,65536,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,2048,0.006666666517655055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,64,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,256,32,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,51200,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,65536,0.046216888560189136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,65536,0.04623200164900886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,256,512,0.0024862223201327855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,65536,0.13364711072709826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,16384,0.006215111249023014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,5120,0.018932445181740653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,51200,0.0373617774910397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,51200,0.037243554989496865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,12288,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,16384,0.015838222371207345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,16384,0.015278221832381355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,10240,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,12288,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,12288,0.013272888958454132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,12288,0.031926221317715116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,8192,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,10240,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,10240,0.011419555379284753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,128,0.007620444728268518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,7168,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,8192,0.010654222634103564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,8192,0.010050666828950247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,6144,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,8192,0.024705777565638225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,7168,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,7168,0.009473777479595607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,5120,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,7168,0.022704000274340313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,6144,0.009343110852771336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,6144,0.008966222405433655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,4096,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,6144,0.020948444803555805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,5120,0.008658666577604082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,5120,0.008053333395057255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,51200,0.10598666800392999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,3584,0.005126222140259213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,5120,0.018883556127548218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,4096,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,3072,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,4096,0.01719466679626041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,3584,0.007909333540333642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,3584,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,2560,0.00498488876554701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,3584,0.015211555692884656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,3072,0.006755555669466655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,10240,0.028135998381508723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,2048,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,3072,0.014846222268210517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,2560,0.007246221933099959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,2560,0.006413333531883027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,1536,0.0035422220826148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,2560,0.013409777647919126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,2048,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,2048,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,1024,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,2048,0.013106666505336761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,1536,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,1536,0.005976000179847081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,768,0.002846222163902389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,1536,0.011353777514563667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,1024,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,1024,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,512,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,4096,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,768,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,1024,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,768,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,16384,0.040362666050593056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,768,0.009913778139485253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,512,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,128,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,512,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,256,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,256,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,3072,0.0074480002125104266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,64,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,256,0.009305777649084726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,32,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,65536,0.008435555630260045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,128,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,128,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,51200,0.007363555332024892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,128,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,16384,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,12288,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,32,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,65536,0.0458826654487186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,51200,0.036962668100992836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,16384,0.015141333142916361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,8192,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,12288,0.01275022245115704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,7168,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,10240,0.011312888728247749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,6144,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,8192,0.010010666317409938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,128,256,0.002163555473089218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,5120,0.004604444321658877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,7168,0.009560888840092553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,4096,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,3584,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,6144,0.00869422240389718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,3072,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,512,0.009339555270142024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,5120,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,2560,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,2048,0.004562666846646203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,4096,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,1536,0.003527111063400904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,3584,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,1024,0.0028248888750871024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,3072,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,128,64,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,768,0.0024862223201327855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,512,0.002206222257680363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,1536,0.005980444451173146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,256,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,1024,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,128,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,768,0.00527555536892679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,64,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,512,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,32,0.00180177783800496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,256,0.004968889057636261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,65536,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,128,0.004581333448489507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,51200,0.008020444048775567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,16384,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,64,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,32,0.004750222381618288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,65536,0.04638488756285774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,51200,0.03702133231692844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,10240,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,8192,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,16384,0.015101333459218344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,7168,0.005042666776312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,12288,0.01255733354224099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,6144,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,10240,0.011367110742463006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,5120,0.004586666822433472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,8192,0.010317333042621613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,2560,0.006653333289755716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,4096,0.004521777646409141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,64,2048,0.006046222315894232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,3584,0.004874666531880696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,6144,0.008678221868144141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,64,10240,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,2560,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,5120,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,2048,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,4096,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,3584,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,1536,0.003540444291300244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,1024,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,3072,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,768,0.002465777720014254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,2560,0.0063573333124319715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,512,0.0022177776942650476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,2048,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,1536,0.005624889085690181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,12288,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,1024,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,128,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,768,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,64,0.0017973333597183228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,32,0.0018035556293196147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,512,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,256,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,128,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,64,0.0047182221379545005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,32,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,80,32,7168,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,3072,0.004581333448489507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,16384,0.3835137685139974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,16384,0.19887644714779326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,12288,0.29034045007493764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,51200,0.5988924238416884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,16384,0.13100977738698324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,80,32,256,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,10240,0.21391467253367105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,51200,1.2095858256022136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,12288,0.1477137804031372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,12288,0.0992444422509935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,10240,0.12723910808563232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,10240,0.09532711240980361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,8192,0.10611377822028266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,51200,0.3885768784417047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,7168,0.09338933229446411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,7168,0.059791107972462974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,5120,0.11212621794806586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,8192,0.07012977865007189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,6144,0.052651554346084595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,6144,0.07828799883524577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,4096,0.09130311012268066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,5120,0.046556442975997925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,5120,0.06581333610746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,3584,0.07897777689827813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,4096,0.054795556598239474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,8192,0.17633867263793945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,4096,0.046431110964881055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,3072,0.06778222322463989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,3584,0.035495112339655556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,2560,0.05678844451904297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,7168,0.15294399526384142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,3072,0.03052977720896403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,2048,0.04717511269781324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,6144,0.1318897803624471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,1536,0.03613955444759793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,2560,0.031303998496797346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,2560,0.0367253323396047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,1024,0.02682755556371477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,2048,0.022617777188618977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,2048,0.03186666634347703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,1536,0.018572444717089336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,768,0.0199626667631997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,1024,0.014789332946141561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,512,0.013740444348918067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,1024,0.03205600049760606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,768,0.013280889226330651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,256,0.00812977800766627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,768,0.024696000748210486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,512,0.012291555603345236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,128,0.006334222025341458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,512,0.018518222702874076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,256,0.010764444039927589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,256,0.012773333324326409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,64,0.004360888981156879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,65536,128,0.010685332947307162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,65536,32,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,3584,0.04840088884035746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,128,0.011095110740926532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,3072,0.042091555065578885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,64,0.011384888655609555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,32,0.01534933348496755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,65536,1536,0.04416444566514757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,16384,0.27497778998480904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,65536,0.6213564342922634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,51200,0.471268441942003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,51200,0.849089781443278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,16384,0.1576817830403646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,12288,0.20729866292741564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,16384,0.10482222504085964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,10240,0.17374755276574028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,12288,0.11850133207109238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,12288,0.08001422219806247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,51200,0.3147786723242866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,8192,0.14066933261023626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,10240,0.0668426685863071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,7168,0.12383288807339138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,8192,0.08443822463353474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,7168,0.04909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,8192,0.05608533488379585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,6144,0.10633689165115356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,10240,0.09965599907769097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,7168,0.07166044579611884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,6144,0.042929778496424355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,65536,1.0846231248643663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,4096,0.07344177696439955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,5120,0.037237332926856145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,6144,0.06276088953018188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,5120,0.05310577816433377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,3584,0.0648133357365926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,3072,0.05703733364741007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,4096,0.043990221288469106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,3584,0.02910311023394267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,3072,0.025935999221271936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,3584,0.0395066671901279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,3072,0.034367998441060386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,65536,0.41125067075093585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,2048,0.0388355553150177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,1536,0.03051911128891839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,2560,0.026714666022194758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,2048,0.026141333911154006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,1536,0.016184889607959323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,2048,0.01942755613062117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,1536,0.03578666514820523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,768,0.016228444046444364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,1024,0.01347555551264021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,5120,0.09079200029373169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,512,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,1024,0.025628444221284654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,768,0.012251555919647217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,768,0.019917334119478863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,256,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,512,0.010605333579911126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,128,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,512,0.015007999208238391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,256,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,256,0.01070577816830741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,64,0.003882666842805015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,2560,0.0475466681851281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,128,0.009712888962692684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,32,0.00416088890698221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,51200,4096,0.03220444586541917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,128,0.009775111244784461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,2560,0.030255999830034044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,64,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,51200,32,0.01311555587583118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,65536,0.319392893049452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,51200,1024,0.02257066633966234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,51200,0.24899554252624512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,16384,0.08146844307581584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,65536,0.2213519944085015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,12288,0.06404088603125678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,65536,0.1490293343861898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,51200,0.1706391175587972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,51200,0.11223199632432725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,10240,0.0528728895717197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,16384,0.041005333264668785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,12288,0.04524711105558607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,8192,0.043975111511018544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,12288,0.03492622243033515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,10240,0.029676444000667993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,10240,0.03908177879121568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,8192,0.02384800049993727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,6144,0.033222221665912204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,8192,0.03390844331847297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,7168,0.02145333256986406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,5120,0.0284551117155287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,7168,0.029530667596393164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,6144,0.019168888529141743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,4096,0.023711999257405598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,5120,0.016787555482652452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,3584,0.020933333370420668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,6144,0.026407111022207472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,5120,0.02271555529700385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,4096,0.015386665860811869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,4096,0.01924977699915568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,3584,0.01419911119672987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,2560,0.015407999356587728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,3072,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,3584,0.017254221770498488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,2048,0.012568888564904531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,3072,0.015434667468070984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,2560,0.01350311107105679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,1536,0.010009778042634329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,2560,0.013786666923099093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,2048,0.010710222025712332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,1024,0.007814221911960179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,1536,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,2048,0.0123857781291008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,768,0.006312888943486744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,1024,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,1536,0.015170666906568738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,7168,0.03844711184501648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,512,0.004652444273233414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,256,0.0029804443733559716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,768,0.00740444411834081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,768,0.010002666877375709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,512,0.007169778148333232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,256,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,512,0.008041777544551426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,64,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,256,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,32,0.0024862223201327855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,16384,128,0.006671999891599019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,3072,0.018205333087179396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,64,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,128,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,65536,0.2472471131218804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,51200,0.1957697735892402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,65536,0.17135111490885416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,16384,0.059897780418395996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,65536,0.12670844131045872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,16384,0.069541335105896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,51200,0.13074755668640137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,12288,0.05382755398750305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,1024,0.011959999799728394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,16384,0.03682400120629205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,10240,0.04451466600100199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,16384,0.04689066608746847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,16384,128,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,51200,0.10424711306889851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,12288,0.03267733256022135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,8192,0.0369777778784434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,12288,0.036093334356943764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,7168,0.03250755535231696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,10240,0.02529688841766781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,8192,0.021607110897699993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,6144,0.028060442871517603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,10240,0.031192888816197712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,7168,0.019662222928471036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,5120,0.024285333024130926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,4096,0.020392888122134738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,6144,0.017976888351970248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,16384,32,0.008096888661384583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,7168,0.023711999257405598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,6144,0.02117777698569828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,3584,0.017794667018784415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,5120,0.018287110659811232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,3072,0.015600888265503777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,4096,0.01556799974706438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,3584,0.013329777452680798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,2560,0.013056000073750814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,3584,0.01427288850148519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,3072,0.012633778154850006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,2048,0.011156444748242697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,2560,0.012491555677519904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,2560,0.011974222130245633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,1536,0.009115555220180089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,2048,0.009934222532643212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,2048,0.010939555863539377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,1024,0.007344000041484833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,1536,0.008512000242869059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,1536,0.013007111019558378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,768,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,8192,0.026671111583709717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,1024,0.010015111002657149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,512,0.0042160000238153665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,768,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,768,0.008596444295512306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,256,0.0032133333798911837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,4096,0.014256889621416727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,512,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,512,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,128,0.002833777831660377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,256,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,64,0.0027022223091787765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,128,0.00647288892004225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,3072,0.01314222233162986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,12288,32,0.00282577777074443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,128,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,64,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,65536,0.22201244036356607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,32,0.00737955586777793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,65536,0.16264266437954372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,51200,0.1743697722752889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,65536,0.12416177325778538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,16384,0.06332177586025663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,1024,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,51200,0.12529955969916448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,16384,0.03507466779814826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,12288,5120,0.0158915552828047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,10240,0.04055111275778876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,16384,0.04485333296987745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,12288,0.029496888319651287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,12288,0.03446844551298354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,8192,0.03410666518741184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,7168,0.029986666308508977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,10240,0.025058666865030926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,10240,0.029870222012201946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,12288,256,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,8192,0.026343110534879897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,7168,0.019000000423855253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,5120,0.022242666946517095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,7168,0.022280888424979314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,6144,0.01723911033736335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,6144,0.019941333267423842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,4096,0.01891377733813392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,5120,0.015434667468070984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,3584,0.016220443778567843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,5120,0.01721244388156467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,4096,0.01388711151149538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,3072,0.014127999544143677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,4096,0.01497333414024777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,3584,0.013056000073750814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,2560,0.012390221986505719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,12288,0.048904889159732394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,3584,0.01385600037044949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,3072,0.012011555333932241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,2048,0.010673777924643623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,3072,0.012855999999576144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,2560,0.011036444041464063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,2560,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,2048,0.010658666491508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,8192,0.022288888692855835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,1024,0.007244444555706448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,6144,0.02600355611907111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,1536,0.011705777711338468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,768,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,1024,0.007344889144102733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,512,0.0041786668201287585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,768,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,1024,0.009226666556464301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,256,0.0033884443756606844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,768,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,512,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,128,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,512,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,256,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,256,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,128,0.0064586665895250105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,51200,0.09610755576027764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,32,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,128,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,64,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,1536,0.009335999687512716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,65536,0.1725768910513984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,2048,0.00908799966176351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,10240,32,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,51200,0.1320835616853502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,65536,0.12055822213490804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,10240,1536,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,65536,0.07008977731068929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,12288,0.036413331826527916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,51200,0.05684266487757365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,51200,0.09267555342780219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,16384,0.02276177704334259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,10240,0.03013422091801961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,16384,0.033858666817347206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,12288,0.018280888597170513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,12288,0.02661422226164076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,10240,0.0162462227874332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,10240,0.023408000667889912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,7168,0.02239911092652215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,10240,64,0.0025457777082920074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,8192,0.014607111612955729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,6144,0.01945244438118405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,8192,0.019646222392718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,5120,0.01664799948533376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,7168,0.014556444353527494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,7168,0.017431111799346078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,6144,0.012520888613329994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,4096,0.014464888307783337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,6144,0.015828443898095023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,5120,0.011393778026103973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,3584,0.01236355553070704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,5120,0.013773333695199756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,3072,0.01109866632355584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,4096,0.012440000143316058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,3584,0.009388444324334463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,2560,0.009486222432719337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,3584,0.011701333026091257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,3072,0.008734222087595198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,16384,0.050048000282711454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,2048,0.008301333420806462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,3072,0.010776888993051318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,2560,0.008800889054934183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,1536,0.006779555645253923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,2560,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,1024,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,2048,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,1536,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,2048,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,768,0.0042151109211974675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,1536,0.009333333207501305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,1024,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,512,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,768,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,256,0.0026319999661710528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,768,0.00703111125363244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,512,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,128,0.0024871110088295406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,256,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,256,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,64,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,4096,0.010523555179437002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,32,0.0025031111306614345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,8192,128,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,128,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,64,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,65536,0.154783116446601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,32,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,51200,0.12424177593655056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,8192,8192,0.025780444343884785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,16384,0.04457866814401415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,51200,0.05473688907093472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,51200,0.09014311101701523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,12288,0.03439022103945414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,1024,0.008096888661384583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,16384,0.021925333473417494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,16384,0.03272711237271627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,10240,0.028500444359249536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,12288,0.01794399983353085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,8192,0.024866667058732774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,8192,512,0.006320000108745363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,10240,0.015776889191733468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,7168,0.02123733361562093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,8192,0.014166222678290473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,10240,0.02204622162712945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,6144,0.01833422150876787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,8192,0.019488889310095046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,7168,0.013077333569526672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,5120,0.015831111205948722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,7168,0.01718577742576599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,4096,0.013724444641007317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,6144,0.015461333923869662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,5120,0.011027555498811932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,3584,0.011767110890812345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,5120,0.013847110999955071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,4096,0.009753777749008602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,65536,0.11664266056484646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,4096,0.012493333054913415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,3584,0.00902844468752543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,2560,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,3584,0.011567999919255575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,3072,0.008485332959228093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,3072,0.010663111176755695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,2048,0.008069333102968004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,2560,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,12288,0.025536000728607178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,2048,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,2048,0.008071111308203803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,1536,0.00693866651919153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,6144,0.012063999970753988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,1536,0.009947555760542551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,1024,0.004853333450025982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,768,0.003951111187537511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,1024,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,1024,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,3072,0.010646222366227044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,512,0.003289777785539627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,768,0.006017777654859755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,256,0.002595555658141772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,512,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,512,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,128,0.0021964444054497615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,256,0.005699555493063397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,2560,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,256,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,64,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,32,0.0022293333378103045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,128,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,7168,1536,0.006559999866618051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,128,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,64,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,32,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,51200,0.09881244765387641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,65536,0.06532266404893664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,16384,0.03715733355946011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,51200,0.052834666437572904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,51200,0.0855519970258077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,7168,768,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,12288,0.03147555390993754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,16384,0.02102311121092902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,16384,0.0314337776766883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,10240,0.026426666312747534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,12288,0.017297777864668105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,8192,0.02295733326011234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,12288,0.024550222688251074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,10240,0.015441776977645027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,7168,0.019609777463806998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,10240,0.02113955550723606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,6144,0.01686311099264357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,8192,0.0188080006175571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,7168,0.012716444830099741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,5120,0.01463644372092353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,7168,0.01663466625743442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,6144,0.0118231111102634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,4096,0.012050666742854647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,6144,0.015090665883488126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,65536,0.11054133044348823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,3584,0.01146755533085929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,5120,0.013535111314720578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,4096,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,3072,0.011353777514563667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,4096,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,7168,65536,0.06761599911583795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,3584,0.008393777741326226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,2560,0.008745777938101027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,3584,0.011286222272449069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,3072,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,2048,0.0075937774446275495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,2560,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,3072,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,1536,0.006555555595291986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,2560,0.009328000247478485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,2048,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,1024,0.004810666872395409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,2048,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,1536,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,768,0.0038737778862317405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,1536,0.008978666530715095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,1024,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,8192,0.013571555415789286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,512,0.0031386667655573953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,1024,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,256,0.002540444541308615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,65536,0.12382845083872478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,128,0.002435555474625693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,512,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,512,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,128,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,256,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,32,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,128,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,64,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,32,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,65536,0.10626133282979329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,51200,0.08736799822913276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,65536,0.06265688604778714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,16384,0.03231466809908549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,51200,0.05069955521159702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,51200,0.08316000302632649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,12288,0.027434666951497395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,768,0.006144000010357962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,16384,0.020276443825827706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,6144,768,0.00665511108107037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,10240,0.023266666465335425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,6144,64,0.002120888895458645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,12288,0.016574222180578444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,6144,5120,0.010703110860453712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,12288,0.02369955513212416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,8192,0.020560888780487906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,10240,0.015036443869272867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,10240,0.020801777640978496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,65536,0.10809155305226643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,8192,0.013109332985348172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,6144,0.015831111205948722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,8192,0.018067555295096505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,7168,0.012326222327020435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,5120,0.013773333695199756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,6144,0.011362666885058085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,4096,0.01034222212102678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,5120,0.010056888891590966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,5120,0.013170666992664337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,3584,0.009318222602208456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,4096,0.011777777638700275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,16384,0.030842665168974135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,3072,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,3584,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,3584,0.010970667004585266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,2560,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,3072,0.007731555236710443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,3072,0.009833777944246927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,7168,0.018294221825069852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,2560,0.007326222128338284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,2560,0.008353778057628209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,1536,0.006163555714819167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,2048,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,7168,0.01623288955953386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,1024,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,1536,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,1536,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,768,0.0041884444653987885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,1024,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,1024,0.007248000138335758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,512,0.0033946666452619764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,768,0.006031111296680238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,768,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,256,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,512,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,512,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,128,0.0023013332651721104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,256,0.00562755556570159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,256,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,64,0.0021768889079491296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,2048,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,5120,32,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,128,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,128,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,65536,0.08662933111190796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,64,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,51200,0.06994577911165026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,2048,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,16384,0.029310223129060533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,65536,0.06600355439715914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,65536,0.0604648921224806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,12288,0.02233244478702545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,51200,0.051240887906816274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,51200,0.04864799976348877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,16384,0.01942399972014957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,5120,4096,0.008738666772842407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,10240,0.01947111056910621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,16384,0.01995822290579478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,8192,0.01682844426896837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,12288,0.016182222300105624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,7168,0.015110222829712762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,10240,0.014463111758232117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,10240,0.014450665977266101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,6144,0.013730666703648038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,7168,0.011736888852384357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,7168,0.011704000333944956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,5120,0.012040889097584618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,6144,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,6144,0.010888000329335531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,4096,0.009405333134863112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,5120,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,5120,0.00998399986161126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,32,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,3584,0.008267555799749162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,4096,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,3072,0.00740622232357661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,5120,6144,0.014883556299739413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,4096,0.009034666750166152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,2560,0.00656088896923595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,3072,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,3072,0.007342221836249034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,2048,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,2560,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,2560,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,1536,0.0047111109726958806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,12288,0.015930665863884818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,8192,0.01276533305644989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,2048,0.006904888898134232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,1024,0.0035111109415690103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,1536,0.00675022229552269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,2048,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,768,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,1536,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,1024,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,512,0.002806222273243798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,1024,0.006322666588756774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,768,0.005990222096443176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,256,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,768,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,512,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,128,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,512,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,256,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,256,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,3584,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,32,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,128,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,64,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,32,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,8192,0.013320888910028668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,65536,0.0633831090397305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,65536,0.05926577912436592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,51200,0.06663911210166083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,16384,0.02553422252337138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,51200,0.04766311248143514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,51200,0.04924266537030538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,12288,0.020041777027977836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,16384,0.019280888968043856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,4096,64,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,4096,128,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,12288,0.0158897770775689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,4096,3584,0.008310221963458592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,12288,0.0155395550860299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,8192,0.015826667348543804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,10240,0.014392000105645923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,65536,0.08370311392678155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,7168,0.014107555150985718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,10240,0.013962666193644205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,8192,0.012447111308574677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,6144,0.013083555632167392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,8192,0.012624888784355588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,7168,0.01107377807299296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,5120,0.011591110792424945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,6144,0.009723555710580613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,7168,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,4096,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,6144,0.010695111420419482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,5120,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,3584,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,5120,0.009711110757456886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,3072,0.007130666739410824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,4096,0.008482666479216682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,2560,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,4096,0.00830488900343577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,3584,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,3072,0.0075902218619982404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,2048,0.00535111129283905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,16384,0.018975999620225694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,3072,0.007277333074145847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,1536,0.004190222256713443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,2560,0.007076444725195567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,2048,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,1024,0.0034968890249729156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,1536,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,2048,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,768,0.0031475555151700974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,1536,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,1024,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,512,0.0027111110587914786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,768,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,1024,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,256,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,512,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,768,0.0058497778243488735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,128,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,512,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,256,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,64,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,32,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,256,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,128,0.00489688871635331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,65536,0.07352355453703138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,64,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3584,32,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,3584,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,51200,0.05851022402445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,16384,0.020804444948832195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,65536,0.06057244539260864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,65536,0.05817955732345581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,2560,0.007518221934636434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,12288,0.017316444052590262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3584,10240,0.017264889346228707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,51200,0.04740088846948412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,16384,0.01904177831278907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,10240,0.01482399966981676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,8192,0.012693333129088083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,12288,0.015134221977657743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,10240,0.014077333940400017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,10240,0.013554666605260638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,8192,0.011651555697123209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,7168,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,8192,0.012437333663304647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,6144,0.011167999770906238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,7168,0.010674667027261523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3584,128,0.005681777579916849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,5120,0.009646221995353699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,7168,0.011195555329322815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,6144,0.009755555954244401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,6144,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,4096,0.00833777752187517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,5120,0.009135111338562435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,3584,0.0075839997993575195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,5120,0.009176889227496253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,4096,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,3072,0.0069022224181228215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,4096,0.008001777860853408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,3584,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,51200,0.04657333427005344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,2560,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,3584,0.007738666401969061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,3072,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,2048,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,16384,0.018256000346607633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,3072,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,2560,0.007381333245171442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,1536,0.004299555387761858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,1024,0.003539555602603488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,2560,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,2048,0.007091555330488417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,768,0.0031831111345026228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,1536,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,2048,0.006325333482689328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,512,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,1536,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,768,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,1024,0.006043555421961679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,768,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,512,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,128,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,512,0.0053191110491752625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,256,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,64,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,256,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,128,0.0058151111006736755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,32,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,128,0.004978666702906291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,65536,0.06272800101174249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,64,0.005130666825506422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,51200,0.04647466540336609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,3072,32,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,65536,0.057209776507483594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,16384,0.018716444571812946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,65536,0.05791288614273071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,12288,0.014800889624489678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,51200,0.045680890480677284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,12288,0.015729778342776828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,16384,0.018805333309703402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,51200,0.045555555158191256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,12288,0.015160888433456421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,16384,0.017888888716697693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,8192,0.011752000285519494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,10240,0.012847111456924014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,12288,0.01475822263293796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,3072,1024,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,7168,0.011053333679835001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,10240,0.013222222526868185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,6144,0.010061333576838175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,8192,0.011196444431940714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,3072,256,0.0024577778660588795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,5120,0.009336888790130615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,8192,0.011798222031858234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,7168,0.010390222072601318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,4096,0.00815288888083564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,6144,0.009661333428488838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,7168,0.010792888700962067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,3584,0.007368000017272101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,5120,0.008990222381220924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,6144,0.00980355590581894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,3072,0.006842666616042455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,4096,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,5120,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,2560,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,3584,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,4096,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,2048,0.005034666508436203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,3072,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,3584,0.007371555599901412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,1536,0.003991999973853429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,2560,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,3072,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,1024,0.003383999897374047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,2048,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,768,0.002887111157178879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,1536,0.006778666542636023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,1024,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,1536,0.0063973334100511335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,512,0.002500444443689452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,1024,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,10240,0.013447110851605734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,256,0.002248888835310936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,768,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,512,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,128,0.002161777681774563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,512,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,256,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,64,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,128,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,128,0.004976888911591636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2560,32,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,65536,0.051072001457214355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,64,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,32,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,51200,0.04036888811323378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,65536,0.055789331595102944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,65536,0.0551493333445655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,2560,0.00676711110605134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,16384,0.016283555163277518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,2048,0.0064026667839950984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,51200,0.044863111442989774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,51200,0.04380977816051907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,16384,0.01835911141501533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,10240,0.011821333732869891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,12288,0.015000889698664347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,12288,0.014447111222479077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,8192,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,10240,0.012657778130637275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2560,256,0.005079999979999331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,10240,0.013069333301650153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,7168,0.009916444619496664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,8192,0.01108977778090371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,8192,0.011549333731333414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,7168,0.010128888818952773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,7168,0.010695111420419482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,5120,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,6144,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,6144,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,4096,0.009289777941173976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,5120,0.008851555486520132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,5120,0.008404444489214156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,3584,0.008632000121805403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,12288,0.013418667018413544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,4096,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2560,768,0.006063111126422882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,4096,0.007784888976150089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,16384,0.017515555024147034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,3584,0.008030222521887885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,2560,0.006684444430801604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,3584,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,3072,0.007620444728268518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,2048,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,2560,0.007323555648326874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,1536,0.005042666776312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,2560,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,2048,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,1024,0.003936000168323517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,1536,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,2048,0.006310222049554189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,768,0.0028844444702068963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,6144,0.010110222631030613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,1536,0.006302222195598815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,1024,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,1024,0.0058515556156635284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,768,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,256,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,512,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,768,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,128,0.0021564445147911706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,512,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,256,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,64,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,32,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,2048,128,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,256,0.005212444398138259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,128,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,64,0.004948444250557157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,32,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,51200,0.030979557169808283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,65536,0.054808888170454234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,65536,0.052991112073262535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,16384,0.013471999930010902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,2048,3072,0.006973333656787872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,16384,0.01753155555990007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,51200,0.04206933246718513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,16384,0.01696266730626424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,10240,0.011815999945004782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,12288,0.013753777576817406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,512,0.0026337777574857077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,8192,0.011023999916182624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,12288,0.014127110441525778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,10240,0.012277333272827996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,7168,0.010405333505736457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,10240,0.012726222475369772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,8192,0.010689777632554373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,6144,0.009909333454238044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,8192,0.010724444356229572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,7168,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,5120,0.008605333666006723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,65536,0.038875556654400296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,7168,0.009768889182143742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,6144,0.009635555247465769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,4096,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,6144,0.009263111485375298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,5120,0.008978666530715095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,3584,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,4096,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,5120,0.008599999878141616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,3072,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,4096,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,51200,0.044255110952589244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,3584,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,2560,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,3584,0.0074888889988263445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,3072,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,2048,0.004290666845109728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,2560,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,3072,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,1536,0.004515555583768421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,2048,0.0069368887278768755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,2560,0.006691555596060223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,1024,0.0032133333798911837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,1536,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,768,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,1536,0.0060159998635451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,1024,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,512,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,1024,0.005786666853560342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,768,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,256,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,768,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,512,0.005688888745175467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,128,0.0018613334331247543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,2048,3072,0.007487110793590546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,64,0.0017813332378864288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,256,0.00519822206762102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1536,128,0.005640888793600931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,32,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,65536,0.028424888849258423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,128,0.00499288903342353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1536,12288,0.013728000223636627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,64,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,51200,0.02292088833120134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,32,0.0052844443255000645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,65536,0.05454666746987236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,16384,0.011546666423479715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,65536,0.05045244428846571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,51200,0.043883555465274386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,16384,0.016480889585283067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,51200,0.04031999905904134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,2048,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,10240,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,16384,0.016454221473799813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,8192,0.009299555586444007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,12288,0.013312000367376538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,7168,0.010487111078368293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,12288,0.012995555169052549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,8192,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,10240,0.011698666546079846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,6144,0.009441778063774109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,8192,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,7168,0.010055999788973067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,5120,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,7168,0.009731555150614845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1536,512,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,6144,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,4096,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,3584,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,5120,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,6144,0.009007111191749573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,3072,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,5120,0.008435555630260045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,3584,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,4096,0.007660444411966536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,2560,0.005193777796294955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,3072,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,2048,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,3072,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,2560,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,1536,0.0047235555118984645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,2560,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,12288,0.010995555255148145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,2048,0.006759111252095964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,2048,0.006331555545330048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,1536,0.00656444455186526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,1536,0.005969777703285217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,1024,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,10240,0.012083555261294046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,1024,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,512,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,768,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,256,0.002465777720014254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,768,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,512,0.005877333382765452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,128,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,512,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,256,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,64,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,256,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,4096,0.008249777886602614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,1024,128,0.005414222263627582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,32,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,128,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,65536,0.022841778066423204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,3584,0.007344000041484833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,64,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,1024,32,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,65536,0.05402222275733948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,16384,0.011038222246699862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,65536,0.04980355501174927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,1024,0.004562666846646203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,51200,0.043331556850009494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,51200,0.03979644510481093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,16384,0.015972443752818637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,1024,768,0.0032088889016045462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,10240,0.008722667064931657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,16384,0.016154666741689045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,12288,0.013154666456911298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,8192,0.008725333544943068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,12288,0.012764444781674279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,7168,0.008349333372380998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,8192,0.010737777584128909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,10240,0.011699555648697747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,6144,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,8192,0.010394666757848527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,7168,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,5120,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,7168,0.009646221995353699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,6144,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,6144,0.008985777695973715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,5120,0.008640888664457533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,51200,0.022491556074884202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,3584,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,4096,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,4096,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,3072,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,3584,0.007873777714040544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,12288,0.00941777808798684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,2560,0.005074666606055366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,3584,0.0074133334888352295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,2048,0.00453955555955569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,3072,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,3072,0.00702311098575592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,2560,0.007206222249401941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,10240,0.012088889049159156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,2048,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,1024,0.00315022220214208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,2048,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,1536,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,1536,0.00598933340774642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,1024,0.005988444305128521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,1024,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,512,0.0024906665914588505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,4096,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,768,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,256,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,768,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,5120,0.008472888833946651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,128,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,512,0.0052879999081293745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,256,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,64,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,128,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,32,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,128,0.005123555660247803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,65536,0.020688888099458482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,64,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,32,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,768,2560,0.006606222026877933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,51200,0.01716711123784383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,65536,0.05268889003329807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,65536,0.049067556858062744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,16384,0.011824888487656912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,768,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,12288,0.009858667022652095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,51200,0.03922577699025472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,16384,0.015837333268589444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,10240,0.008072000410821702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,16384,0.015160888433456421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,12288,0.013176889055305056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,8192,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,12288,0.01291822228166792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,512,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,10240,0.012114666402339935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,7168,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,768,256,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,10240,0.011398221883508893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,8192,0.01072355525361167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,6144,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,5120,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,7168,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,6144,0.009639110830095079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,7168,0.009711999860074785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,4096,0.00657155571712388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,5120,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,3584,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,5120,0.008415111237102086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,768,1536,0.0038515557017591265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,4096,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,3072,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,4096,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,3584,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,2560,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,3072,0.007624000310897827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,3072,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,51200,0.038667556312349104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,2048,0.005323555734422471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,2560,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,1536,0.003907555507289039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,2048,0.00675999994079272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,2048,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,1536,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,1536,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,8192,0.010271111296282874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,768,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,1024,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,1024,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,512,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,6144,0.009006222089131674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,768,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,256,0.0022035555707083806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,512,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,512,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,128,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,3584,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,256,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,64,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,256,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,32,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,512,128,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,65536,0.017873777283562552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,128,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,2560,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,51200,0.015875554747051664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,64,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,32,0.004951111144489712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,65536,0.04657599992222256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,512,1024,0.0032186667538351486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,51200,0.03735911183887058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,65536,0.04714222086800469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,12288,0.007678222325113084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,16384,0.015788444214397006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,51200,0.03779200050565932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,10240,0.006205333189831839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,16384,0.015140444040298462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,12288,0.01309511148267322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,8192,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,12288,0.012736000120639801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,10240,0.011875555747085147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,7168,0.006650666809744305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,10240,0.011571555501884885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,8192,0.010676444404655032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,6144,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,8192,0.010203555226325989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,7168,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,5120,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,7168,0.00943022221326828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,6144,0.00941777808798684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,4096,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,3584,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,6144,0.009031111167536842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,5120,0.008943110704421997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,4096,0.008150222400824228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,3072,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,5120,0.008343111309740279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,2560,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,3584,0.007721777591440413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,4096,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,2048,0.004200889004601372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,3072,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,2560,0.007256000406212277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,16384,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,3072,0.007009777757856581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,1536,0.003552888830502828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,1024,0.003072888900836309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,2048,0.006246222390068903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,1536,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,768,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,1536,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,1024,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,1024,0.005292444593376583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,512,0.002487999904486868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,768,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,256,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,512,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,512,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,128,0.0018213333355055915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,256,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,256,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,64,0.001810666587617662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,128,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,256,32,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,128,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,65536,0.011332444846630096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,3584,0.007228444847795699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,64,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,32,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,256,2048,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,65536,0.04607289036115011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,65536,0.045147554741965405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,16384,0.0069146665434042616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,51200,0.03759555684195624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,51200,0.03631466627120972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,12288,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,16384,0.015861334072219003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,16384,0.014883556299739413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,512,768,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,10240,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,12288,0.013357333011097379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,8192,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,12288,0.012431999875439538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,10240,0.011720000041855706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,7168,0.005305777821275923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,10240,0.011114666859308878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,8192,0.010688888529936472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,6144,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,8192,0.01033688916100396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,7168,0.01000711073478063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,5120,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,7168,0.0094373333785269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,6144,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,4096,0.005772444523043103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,51200,0.01239466667175293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,2560,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,5120,0.008700444466537898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,3584,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,5120,0.008107555409272512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,4096,0.008242666721343994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,3072,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,4096,0.00758844448460473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,3584,0.007676444119877285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,2560,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,3584,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,3072,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,2048,0.0041911109454101985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,2560,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,256,768,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,1536,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,2560,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,1024,0.0029031110720502008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,1536,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,1536,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,768,0.002846222163902389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,1024,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,1024,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,512,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,768,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,768,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,256,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,6144,0.008802666432327693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,512,0.004984000076850255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,128,0.0018844445132546956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,256,0.005375111268626318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,256,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,64,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,128,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,128,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,65536,0.012823999755912356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,3072,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,51200,0.011378666592968835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,2048,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,64,0.005330666485759947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,32,0.004592000196377437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,128,2048,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,12288,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,10240,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,65536,0.04519911275969612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,8192,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,51200,0.03663644525739882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,7168,0.0052844443255000645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,12288,0.012422222230169507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,10240,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,8192,0.010044444766309526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,6144,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,5120,0.0052977779673205475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,512,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,4096,0.005115555392371284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,3584,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,7168,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,3072,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,6144,0.008709333009190029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,5120,0.008010666403505538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,2560,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,4096,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,2048,0.004294222427739037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,65,128,128,0.005374222166008419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,1536,0.0035377778112888336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,3584,0.00700800038046307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,1024,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,768,0.002642666714058982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,3072,0.006999111009968652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,512,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,2560,0.0063893331421746155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,256,0.0021280000607172647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,16384,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,1024,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,128,0.0018613334331247543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,64,0.001791999985774358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,768,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,512,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,64,32,0.0018115554832749898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,256,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,65536,0.0148044443792767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,128,0.004569777597983678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,51200,0.01274222218328052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,64,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,16384,0.015170666906568738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,16384,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,12288,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,32,0.005156444592608346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,10240,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,65536,0.044975111881891884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,51200,0.03629511263635423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,16384,0.015138667490747241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,7168,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,12288,0.012402666939629449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,6144,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,10240,0.01148799972401725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,5120,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,8192,0.009919111099508073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,4096,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,7168,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,3584,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,2048,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,64,1536,0.00574222207069397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,3072,0.004589333302444882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,5120,0.008303999900817871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,2560,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,2048,0.004578666554556953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,4096,0.007357333269384172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,1536,0.003539555602603488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,3584,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,1024,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,3072,0.006879111131032308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,2560,0.006654222392373615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,768,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,2048,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,512,0.002461333241727617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,256,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,1536,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,128,0.001826666709449556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,1024,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,64,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,768,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,512,0.005003555367390315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,32,0.001783999924858411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,65,32,8192,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,256,0.005052444421582752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,128,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,64,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,32,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,65,32,6144,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,16384,0.3660453425513373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,16384,0.1676079961988661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,51200,0.4967484474182129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,12288,0.27415643797980416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,16384,0.13674133353763157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,12288,0.12491999732123481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,51200,1.1295928955078125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,10240,0.22828533914354113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,10240,0.10593689150280422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,12288,0.10120089186562432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,51200,0.4304204516940647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,10240,0.0868320001496209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,8192,0.1828835540347629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,8192,0.08880800008773804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,7168,0.16003555721706814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,8192,0.07106933328840467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,16384,0.28425778283013237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,7168,0.07560533285140991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,12288,0.217285328441196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,6144,0.1390213304095798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,7168,0.06717689169777764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,5120,0.12193600336710612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,8192,0.14482666386498347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,6144,0.058147556251949735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,10240,0.1772702270083957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,6144,0.06635377804438274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,4096,0.09639110830095078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,5120,0.05599821938408745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,5120,0.04841777682304382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,4096,0.04652888907326592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,7168,0.12432977888319229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,6144,0.11279466417100693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,3584,0.08402044243282741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,3072,0.07304355833265516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,4096,0.03969066672854953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,3584,0.03484088844723172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,5120,0.0938746664259169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,4096,0.07987022399902344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,3584,0.04163288738992479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,3072,0.036632001399993896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,2560,0.06153421931796604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,3072,0.030903998348448012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,2048,0.05039911137686836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,3072,0.0646631121635437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,2560,0.029923554923799302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,2560,0.03167466653717889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,1536,0.03962488969167074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,2048,0.023481777972645227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,2048,0.027109333210521277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,2560,0.055832001898023814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,1024,0.029042667812771265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,2048,0.04814133379194471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,1536,0.02203911046187083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,1536,0.018819555640220642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,768,0.021219554874632094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,1024,0.015423110789722867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,1536,0.04206933246718513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,1024,0.016410667035314772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,512,0.01481066644191742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,1024,0.03386133246951633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,768,0.013741333451535968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,768,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,256,0.008316444853941599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,512,0.012379555238617791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,512,0.02466222147146861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,3584,0.07181866963704427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,128,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,256,0.00925688859489229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,256,0.010763555765151978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,256,0.02017333275742001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,64,0.003850666599141227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,128,0.010681777364677854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,128,0.008316444853941599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,65536,32,0.003925333420435588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,128,0.016153777639071148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,64,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,32,0.012413333687517377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,51200,0.8842408921983508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,65536,512,0.0107004443804423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,65536,0.5250550905863444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,65536,1.1078071594238281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,51200,0.40271110004848903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,51200,0.8655333518981934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,16384,0.2829848925272624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,16384,0.1348231103685167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,65536,0.40719466739230686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,768,0.029104888439178467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,12288,0.21114577187432182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,51200,0.33277689086066353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,12288,0.10136533445782131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,16384,0.10782933235168457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,10240,0.17897244294484457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,12288,0.09062666363186306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,10240,0.08594844738642375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,8192,0.14395911163753933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,10240,0.07648177941640218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,16384,0.22993156645033094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,12288,0.17059644063313803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,7168,0.12510311603546143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,10240,0.14533155494266087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,7168,0.061795552571614586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,7168,0.0591191119617886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,6144,0.10813866721259223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,8192,0.11880977948506673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,7168,0.10587555832333034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,6144,0.05413866705364651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,5120,0.09139821926752727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,6144,0.045477333996031016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,8192,0.07234666744867961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,8192,0.05756088760164049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,5120,0.04613599843449063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,4096,0.07376622491412692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,5120,0.03875022133191427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,6144,0.0936559968524509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,3584,0.0655404461754693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,4096,0.03833511140611436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,4096,0.032893333170149065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,3072,0.05711911122004191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,3584,0.03444622291458978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,3584,0.02951644526587592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,4096,0.06800978051291572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,3584,0.06105066670311821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,2560,0.048511998520957104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,3072,0.03052622079849243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,3072,0.026198221577538386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,2048,0.03949688871701559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,3072,0.05584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,2560,0.024720888998773363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,2560,0.048728889889187284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,2048,0.022674666510687932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,2048,0.019306666321224637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,1536,0.031179553932613794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,2048,0.04277866747644213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,1536,0.016167999969588388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,1024,0.02223555578125848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,1536,0.01819644371668498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,5120,0.08044622341791789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,1536,0.03660355673895942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,768,0.016470222009552848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,1024,0.013224889006879596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,1024,0.01351466692156262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,65536,0.9372889200846354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,512,0.011424000064531961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,1024,0.029476443926493328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,768,0.012109333442317115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,768,0.011456888582971362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,256,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,768,0.024873778223991394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,512,0.011160888605647616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,512,0.010512000156773461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,128,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,51200,0.700218677520752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,256,0.007816000117195977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,512,0.02176888949341244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,64,0.0038524443904558816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,256,0.018609777092933655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,128,0.009707555174827576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,51200,128,0.01418133411142561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,64,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,128,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,32,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,65536,0.3523084322611491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,65536,0.1784062253104316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,65536,0.14824177159203425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,51200,0.2750728925069173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,51200,2560,0.026737777723206416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,51200,256,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,16384,0.08294222089979383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,51200,32,0.0041786668201287585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,16384,0.04929155442449781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,16384,0.04172177778349983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,12288,0.06433600187301636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,16384,0.0581173300743103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,51200,0.12029777632819282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,10240,0.052574220630857676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,12288,0.03758577836884393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,12288,0.03596533338228861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,12288,0.04503200120396084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,65536,0.18655021985371908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,8192,0.0453324450386895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,10240,0.029383109675513372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,10240,0.03255911005867852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,51200,0.13548178142971462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,7168,0.03802311089303758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,51200,0.14304888248443604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,10240,0.039226667748557195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,8192,0.02423911127779219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,6144,0.033215999603271484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,8192,0.028174221515655518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,7168,0.024192000428835552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,5120,0.028416888581381902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,7168,0.02444800072246128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,8192,0.03497066762712266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,6144,0.01994044416480594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,4096,0.024344889654053584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,7168,0.03112444281578064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,6144,0.022321777211295232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,3584,0.020616888999938965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,6144,0.02840977907180786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,5120,0.018966222802797954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,4096,0.015374221735530429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,5120,0.025253333979182776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,3072,0.018202667435010273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,4096,0.015847111741701763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,3584,0.014441778262456259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,4096,0.022514666120211285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,2560,0.0151582227812873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,3584,0.014631110760900708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,3072,0.013389333254761167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,3584,0.020285333196322124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,2048,0.01238222254647149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,3072,0.013478221992651621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,3072,0.018604444132910836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,1536,0.009679111341635386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,2048,0.010946667028797997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,2560,0.01788444485929277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,2048,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,1024,0.007701333198282454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,2048,0.016231111354298063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,1536,0.00962933318482505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,1536,0.009666666388511658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,768,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,1536,0.015113777584499784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,1024,0.013295999831623502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,768,0.006503110958470239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,5120,0.017320889565679762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,512,0.004552000098758274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,768,0.007603555917739868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,768,0.011371555427710215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,256,0.0031742221779293488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,512,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,512,0.007104000283612146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,512,0.009683555199040307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,128,0.002846222163902389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,256,0.006653333289755716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,256,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,2560,0.012033777932325998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,64,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,128,0.0068604445291890045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,128,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,16384,32,0.00282222218811512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,128,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,64,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,1024,0.007840000092983246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,32,0.006650666809744305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,65536,0.24776977962917754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,65536,0.15412978331247965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,51200,0.19450933403438994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,65536,0.13063822852240667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,51200,0.11838311619228786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,16384,0.07025777631335788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,2560,0.012749333348539142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,51200,0.10386933220757379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,16384,0.04299999939070808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,256,0.009541333549552495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,12288,0.04976800084114075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,16384,0.03772266705830892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,16384,0.048659556441836886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,16384,1024,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,10240,0.040369778871536255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,12288,0.0329751107427809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,12288,0.03122133347723219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,65536,0.14897067017025417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,8192,0.03460266523891025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,12288,0.038498666551378034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,51200,0.11790577570597331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,10240,0.025630222426520452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,10240,0.02865333358446757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,7168,0.029422223567962646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,10240,0.03421422176890903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,8192,0.021945777866575453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,8192,0.02535022298494975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,6144,0.02583911187118954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,7168,0.019907555646366544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,7168,0.02183466653029124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,5120,0.022309333086013794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,6144,0.018225777480337355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,7168,0.027612444427278306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,6144,0.019514666663275827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,4096,0.01903377804491255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,5120,0.016145777371194627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,6144,0.02535377773973677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,5120,0.016530666086408827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,3584,0.016169778174824186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,5120,0.022219555245505437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,4096,0.014473777678277759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,4096,0.014553777045673795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,3072,0.014993778533405729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,4096,0.01961688862906562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,3584,0.01349688900841607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,3584,0.013352889153692456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,2560,0.012503110700183444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,3584,0.018630221486091614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,3072,0.013348444468445249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,2048,0.010879110958841113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,3072,0.012401777837011548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,3072,0.01717155509524875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,2560,0.01199288914601008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,2560,0.011337777806652917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,1536,0.008624888956546783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,2560,0.016226665841208566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,2048,0.010059555371602377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,2048,0.010090666512648264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,1024,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,2048,0.015651555524932016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,1536,0.0085235552655326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,8192,0.03074399961365594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,1536,0.008791999684439765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,768,0.0058355554938316345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,1024,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,1536,0.014246222045686511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,512,0.0041866666740841335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,1024,0.006416888700591193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,768,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,1024,0.012424888710180918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,256,0.0028844444702068963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,512,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,768,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,128,0.002537777854336633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,768,0.011046222514576383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,256,0.0063075555695427795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,512,0.009668444593747457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,64,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,256,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,12288,32,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,128,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,256,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,128,0.005290666802061929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,128,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,65536,0.2205831077363756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,64,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,51200,0.17208533816867402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,65536,0.12560533152686224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,16384,0.05257777704132927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,51200,0.11190577348073323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,51200,0.09980889161427815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,16384,0.04053333401679993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,12288,0.041548444165123835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,512,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,16384,0.035644445154401995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,51200,0.11024977763493855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,65536,0.13771466414133707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,12288,0.031522668070263334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,10240,0.03427199853791131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,12288,0.036469333701663546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,12288,32,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,8192,0.029206222958034937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,65536,0.14309333430396184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,10240,0.03236711025238037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,8192,0.023419555690553453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,7168,0.025072889195548162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,8192,0.021120000216695998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,8192,0.029021332661310833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,7168,0.02015911042690277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,7168,0.020453333854675293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,6144,0.021810667382346258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,16384,0.044728888405693897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,6144,0.017620444297790527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,12288,0.028532445430755615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,6144,0.018256000346607633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,5120,0.01883288886811998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,6144,0.02427911096149021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,4096,0.016454221473799813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,5120,0.015836444165971544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,5120,0.015965332587560017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,10240,0.024116444918844435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,3584,0.013544888959990608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,5120,0.021586666504542034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,4096,0.01407999959256914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,4096,0.014011555247836642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,3072,0.012180444267061023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,4096,0.01905155513021681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,3584,0.01293688929743237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,3584,0.013079110946920184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,2560,0.01033688916100396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,3584,0.018557333283954196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,3072,0.011986666255527072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,3072,0.012061333490742577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,2048,0.008781332936551835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,3072,0.017261332935757108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,2560,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,2560,0.011015111373530494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,1536,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,2560,0.016395555602179635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,2048,0.008973333570692275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,2048,0.010019555687904358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,1024,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,1536,0.008356444537639618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,1536,0.007733333441946242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,2048,0.015000889698664347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,768,0.004230222354332606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,1536,0.013927999469969006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,1024,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,10240,0.02715377840730879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,1024,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,512,0.0032177778581778207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,1024,0.011355555719799466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,768,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,768,0.005980444451173146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,256,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,512,0.00693866651919153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,768,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,128,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,256,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,512,0.01020088874631458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,256,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,64,0.002510222295920054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,256,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,128,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,128,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,7168,0.02640977833006117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,64,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,32,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,65536,0.18067999680836996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,51200,0.14094489150577122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,65536,0.09457244475682576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,65536,0.07071822219424777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,51200,0.07330755392710368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,16384,0.049141334162818066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,51200,0.05723110834757487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,16384,0.028353777196672227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,10240,512,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,16384,0.023031110564867657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,12288,0.03269955515861511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,16384,0.04367022381888496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,128,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,65536,0.1361911164389716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,10240,0.027365333504146997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,51200,0.10878755648930867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,12288,0.022037333912319605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,8192,0.02346666653951009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,12288,0.018626666731304593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,10240,0.016300444801648457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,12288,0.03586488962173462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,7168,0.020242666204770405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,8192,0.014470222923490735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,8192,0.01647911138004727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,6144,0.017856000198258292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,8192,0.027684445182482403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,7168,0.014821334017647637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,7168,0.013889777991506787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,5120,0.01534577707449595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,7168,0.025841777523358662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,6144,0.012831110921170978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,6144,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,4096,0.013004444539546967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,6144,0.023249778482649062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,5120,0.01146044416560067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,5120,0.020830222302012973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,4096,0.010753778119881948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,10240,32,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,4096,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,3584,0.012352888782819113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,4096,0.01902222302224901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,3072,0.010687111152542962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,3584,0.009552000297440423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,10240,0.019191111127535503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,3584,0.017614222235149808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,10240,0.031868444548712835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,2560,0.00925155563486947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,3072,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,3072,0.009712888962692684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,2048,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,3072,0.016525333126386006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,2560,0.008619555168681675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,2560,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,1536,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,2560,0.01566844516330295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,2048,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,2048,0.01495911180973053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,5120,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,1024,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,1536,0.0070293330483966405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,1536,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,768,0.003882666842805015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,1536,0.012715555727481842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,1024,0.00656355544924736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,1024,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,512,0.0030906666070222855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,1024,0.011086222198274402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,768,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,768,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,256,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,3584,0.010208889014191097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,768,0.010332444475756751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,512,0.006061333335108227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,128,0.002532444480392668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,512,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,256,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,64,0.002491555487116178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,256,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,128,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,8192,32,0.0026782221264309357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,256,0.009683555199040307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,128,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,128,0.007840000092983246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,65536,0.1545528835720486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,2048,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,64,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,8192,32,0.005719999886221356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,51200,0.12362933158874512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,65536,0.08631466494666205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,16384,0.04456444581349691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,65536,0.06754310925801595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,51200,0.06696089108784993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,51200,0.05453155438105265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,12288,0.03408888975779215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,16384,0.022057778305477563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,16384,0.026296888788541157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,10240,0.026226666238572862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,12288,0.01809244520134396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,12288,0.020575111111005146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,65536,0.13538755310906306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,51200,0.10838933123482598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,8192,0.02261688808600108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,12288,0.03594400154219733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,512,0.009813333551088968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,10240,0.017865777015686035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,10240,0.01571111049917009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,7168,0.019319110446506076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,8192,0.013893332746293811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,8192,0.015125332607163323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,10240,0.031846221950319074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,6144,0.016771554946899414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,7168,0.013891556196742587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,8192,0.027040888865788777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,5120,0.014451555079884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,7168,0.025436444414986506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,6144,0.0120328888297081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,4096,0.012367111113336353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,6144,0.012843555874294706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,6144,0.023072888453801472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,5120,0.011057777537239922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,5120,0.011531555818186866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,3584,0.010656889114114972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,5120,0.021300445000330608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,4096,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,4096,0.010431999961535135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,3072,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,4096,0.018790221876568265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,3584,0.009007111191749573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,3584,0.009942221972677443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,2560,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,3584,0.017539555827776592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,3072,0.008284444610277811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,16384,0.04316177633073595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,2048,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,3072,0.00927022182279163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,3072,0.016926222377353244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,2560,0.007751999629868402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,2560,0.008600888980759515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,2560,0.016164445214801364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,2048,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,2048,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,1024,0.004782222211360931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,7168,0.013094222380055321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,2048,0.014820444915029736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,1536,0.006806222101052602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,1536,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,768,0.0038088887102074097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,1024,0.006314666734801398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,1536,0.012764444781674279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,1024,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,512,0.0031928889867332247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,1024,0.01069155583779017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,768,0.006219555520349079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,768,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,256,0.002704888789190186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,768,0.010371555884679159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,512,0.006120000034570694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,128,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,512,0.010017777482668558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,256,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,256,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,64,0.00240355564488305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,128,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,32,0.0024986666523747975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,128,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,128,0.007702222300900354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,7168,1536,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,64,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,65536,0.12477689319186741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,32,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,51200,0.09820799695120917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,65536,0.06504088640213013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,65536,0.07996177673339844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,16384,0.037576887342664934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,51200,0.060580445660485156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,51200,0.05282577872276306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,16384,0.024334222078323364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,7168,512,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,12288,0.03071999881002638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,65536,0.13532443841298422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,16384,0.020959999826219346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,16384,0.043235556946860425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,256,0.009310222334331935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,12288,0.017183999220530193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,12288,0.018764444523387484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,8192,0.019321777754359774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,10240,0.016511110795868766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,12288,0.03533333208825853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,10240,0.015442666080262927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,7168,0.016391999191708036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,10240,0.031195554468366835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,8192,0.01366488883892695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,6144,0.015136889285511441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,8192,0.014945778581831189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,7168,0.012695999609099494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,7168,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,5120,0.013337777720557319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,7168,0.02458222210407257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,6144,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,6144,0.01236533373594284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,4096,0.011383111278216044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,6144,0.022503111097547743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,5120,0.010603555374675326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,5120,0.011017777853541903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,3584,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,4096,0.009104000197516548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,5120,0.020284444093704224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,10240,0.026603556341595117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,4096,0.009920888476901585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,51200,0.10807289017571343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,3072,0.008568888737095727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,4096,0.018622222873899672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,3584,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,2560,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,3584,0.009356444080670675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,3072,0.008186666501892937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,2048,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,3584,0.017675555414623685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,3072,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,2560,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,2560,0.007373332977294922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,1536,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,2048,0.007127111156781514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,2560,0.01475288967291514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,2048,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,1024,0.0038648889296584656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,2048,0.014697778556081982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,1536,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,1536,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,1024,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,1536,0.011736888852384357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,1024,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,512,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,1024,0.011008888483047485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,768,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,768,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,768,0.010658666491508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,512,0.005285333428117964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,512,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,256,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,512,0.009367999931176504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,128,0.002163555473089218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,256,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,256,0.009032888544930352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,3072,0.016007999579111736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,64,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,128,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,128,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,32,0.0021297778520319196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,128,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,8192,0.026526222626368206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,6144,768,0.0032471112079090546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,64,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,65536,0.11509954929351807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,6144,32,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,51200,0.09180266989601983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,65536,0.07208444674809773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,65536,0.06266044245825873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,16384,0.032045334577560425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,51200,0.0548257761531406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,51200,0.0506897767384847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,12288,0.02749600013097127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,16384,0.022392888863881428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,16384,0.020262221495310467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,65536,0.1351955599255032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,256,0.005886222339338726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,10240,0.023362666368484497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,16384,0.04304000073009067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,12288,0.01687111126052009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,12288,0.01697688963678148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,10240,0.014871999621391296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,8192,0.0192248887485928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,12288,0.034985777404573225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,7168,0.017647999856207106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,10240,0.029932445949978296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,8192,0.013436444103717804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,8192,0.014126222994592456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,6144,0.01554666625128852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,7168,0.012415111064910889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,8192,0.026760889424218073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,7168,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,5120,0.012133333418104382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,7168,0.024064888556798298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,6144,0.011640888949235281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,6144,0.022262222237057153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,4096,0.010384000009960597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,5120,0.010407110883129967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,5120,0.020084444019529555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,4096,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,3584,0.00944977750380834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,51200,0.10772266652848984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,4096,0.009033777647548253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,3072,0.008421333299742805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,4096,0.018429333964983623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,3584,0.008007110820876228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,3584,0.008740444150235917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,10240,0.015146666102939181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,2560,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,3584,0.017092444830470614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,3072,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,2048,0.006010666489601135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,2560,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,3072,0.016494222813182406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,2560,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,1536,0.004821333206362194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,2560,0.014807111687130399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,2048,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,2048,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,6144,0.011191111471917896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,2048,0.013111111190583972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,1536,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,1536,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,5120,0.009909333454238044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,1536,0.012423111332787408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,1024,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,1024,0.01070577816830741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,768,0.006183111005359226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,512,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,768,0.005483555710977978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,768,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,256,0.0024906665914588505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,512,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,512,0.005294222384691238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,3072,0.008045333127180735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,512,0.009405333134863112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,128,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,256,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,256,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,64,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,256,0.00902133352226681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,32,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,128,0.004967999955018361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,128,0.00564177789621883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,1024,0.003524444583389494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,128,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,65536,0.0856284432941013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,64,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,32,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,5120,768,0.0031644445326593188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,51200,0.06981155607435438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,65536,0.06555377774768405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,5120,1024,0.005727111051479976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,16384,0.02752799954679277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,65536,0.0603440006573995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,51200,0.048567109637790255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,51200,0.050322665108574756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,16384,0.0195777780479855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,12288,0.02127200033929613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,16384,0.019719110594855416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,65536,0.1347422202428182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,10240,0.017817777064111497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,16384,0.04233066572083367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,51200,0.10757955577638413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,12288,0.016146666473812528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,8192,0.015098666151364645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,12288,0.015951999359660678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,10240,0.014466666513019137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,12288,0.03374844458368089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,7168,0.013075555364290873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,10240,0.014087110757827759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,8192,0.012648888760142855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,10240,0.030185778935750324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,6144,0.011560888753996955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,8192,0.025990222891171772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,7168,0.011695110963450538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,5120,0.010212444596820408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,7168,0.02446311049991184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,6144,0.010900444454616971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,6144,0.011058666639857821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,4096,0.009447111023796929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,6144,0.021976888179779053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,5120,0.009903999666372934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,5120,0.020370667179425556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,4096,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,3584,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,4096,0.017889777819315594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,3072,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,3584,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,3584,0.007860444486141205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,3584,0.017530666457282174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,2560,0.006486222147941589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,3072,0.007672000262472365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,3072,0.0075911109646161394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,8192,0.013065777719020844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,3072,0.016044444508022733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,2048,0.005537777725193236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,2560,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,2560,0.013790222505728403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,1536,0.004201777693298128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,2048,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,2048,0.013080000049538083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,5120,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,1024,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,1536,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,1536,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,4096,0.008720888859695857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,768,0.0029786665820413162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,1536,0.012206222448084088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,1024,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,1024,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,512,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,1024,0.010709332923094431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,768,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,768,0.005538666827811136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,256,0.002424000038041009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,768,0.01072177787621816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,512,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,512,0.009750222166379293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,256,0.005743999862008625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,256,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,128,0.0021182222084866627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,7168,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,64,0.002157333410448498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,128,0.005667555663320754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,4096,32,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,128,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,128,0.007334222396214803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,64,0.005032000028424793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,65536,0.07554222477806939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,32,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,51200,0.05927733580271403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,65536,0.05918755796220568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,65536,0.06249155600865682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,65536,0.13490577538808188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,51200,0.04760622315936618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,51200,0.04859911070929634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,16384,0.024350222614076402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,512,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,12288,0.019395555059115093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,51200,0.10759822527567546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,2560,0.007256000406212277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,16384,0.01886488828394148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,256,0.009683555199040307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,10240,0.016376000311639573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,12288,0.015864888827006023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,12288,0.015279110934999255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,4096,2048,0.006254222244024277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,8192,0.014455111490355598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,10240,0.014362666341993542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,10240,0.014293332894643148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,10240,0.030007110701666936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,7168,0.01237511138121287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,8192,0.012736000120639801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,8192,0.02637333340115017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,7168,0.011620444556077322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,6144,0.011042666931947073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,7168,0.011345778074529437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,7168,0.023702222439977858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,16384,0.019311111834314134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,6144,0.010676444404655032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,5120,0.009895111123720804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,6144,0.02223199937078688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,16384,0.041778667105568774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,4096,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,5120,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,12288,0.03352800011634827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,5120,0.019917334119478863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,3584,0.0075128889746136135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,4096,0.00851111114025116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,4096,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,8192,0.01239999963177575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,3072,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,4096,0.018405333161354065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,3584,0.008289777570300633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,3584,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,2560,0.006515555497672822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,3584,0.01664177742269304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,3072,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,3072,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,2048,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,3072,0.015171556009186639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,2560,0.007592000067234039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,2560,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,1536,0.004271999829345279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,6144,0.010168000227875179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,2560,0.014149333039919535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,2048,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,1024,0.0031928889867332247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,2048,0.00630044440428416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,2048,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,1536,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,1536,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,768,0.0028693332440323303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,5120,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,1536,0.011709333293967776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,1024,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,512,0.002576888849337896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,1024,0.011212444139851464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,768,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,256,0.002277333289384842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,512,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,768,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,128,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,512,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,512,0.009679111341635386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,256,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,64,0.0018622221218215094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,256,0.009415110780133141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,128,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3584,32,0.0019288888821999233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,128,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,65536,0.06662044260236952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,128,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,64,0.005022222383154763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,51200,0.0539902216858334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,32,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,65536,0.0603297750155131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,65536,0.05812355544832018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,16384,0.020376889242066275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,65536,0.13485066095987955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,51200,0.04741155438952976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,1024,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,12288,0.01722400055991279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,51200,0.10761688815222846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,768,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,16384,0.042040887806150645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,12288,0.014986667368147107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,10240,0.014473777678277759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3584,256,0.005113777601056629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,12288,0.033477332856920027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,8192,0.012445333103338877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,10240,0.014096000128322177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,10240,0.013457777599493662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,10240,0.030214223596784804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,7168,0.011810666984981961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,8192,0.011749332977665795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,8192,0.012432888978057437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,8192,0.02586311101913452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,7168,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,6144,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,7168,0.011085333095656501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,51200,0.04656889041264852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,7168,0.02402755618095398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,5120,0.009461333354314169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,16384,0.018374222848150466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,6144,0.010255110760529837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,16384,0.018949333164427016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,4096,0.008371555142932469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,6144,0.022109333011839125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,5120,0.009129777550697327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,12288,0.015837333268589444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,3584,0.007532444265153673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,5120,0.019888889458444383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,4096,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,4096,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,3072,0.006677333265542984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,4096,0.017494221528371174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,3584,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,3584,0.007335111498832703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,2560,0.0057760001056724125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,3584,0.01581333412064446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,3072,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,2560,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,2560,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,2048,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,2560,0.013732444081041547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,6144,0.009678222239017487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,2048,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,1536,0.004229333251714706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,2048,0.01275466630856196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,5120,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,1024,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,1536,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,1536,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,1536,0.012418666647540199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,768,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,1024,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,1024,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,1024,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,512,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,3072,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,768,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,768,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,256,0.002463999928699599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,3072,0.01516711049609714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,512,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,512,0.00999111102686988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,256,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,256,0.005054222212897407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,2048,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,256,0.009663999908500248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,64,0.0021182222084866627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,32,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,128,0.005669333454635408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,128,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,65536,0.057125336594051786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,128,0.007268444531493717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,64,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,3072,32,0.005295111073387994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,51200,0.045949333243899874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,65536,0.05809511078728569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,65536,0.05698666969935099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,16384,0.021346666746669348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,768,0.00600622221827507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,51200,0.045547554890314736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,65536,0.13446844948662653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,51200,0.04585422078768412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,12288,0.015159999330838522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,512,0.005706666658322017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,51200,0.10730844736099243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,16384,0.017860444055663217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,16384,0.018917333748605516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,10240,0.013430222041077085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,12288,0.015585778488053216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,16384,0.041175110472573176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,12288,0.014546665880415173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,8192,0.01202488856183158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,10240,0.013080000049538083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,12288,0.03312711252106561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,7168,0.01221155540810691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,10240,0.013174222575293647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,10240,0.029245333539115057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,6144,0.011391110718250275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,8192,0.011711999773979187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,7168,0.01037688884470198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,8192,0.026142223013771906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,7168,0.010693333215183683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,5120,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,7168,0.023669333921538457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,6144,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,6144,0.00996800015370051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,4096,0.008162666526105668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,6144,0.021617778473430212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,5120,0.009036444127559662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,3584,0.007234666910436418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,4096,0.00832177781396442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,4096,0.01717155509524875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,3584,0.007381333245171442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,3072,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,3584,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,3584,0.015871110889646742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,3072,128,0.0021964444054497615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,2560,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,3072,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,3072,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,8192,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,3072,0.014654222461912366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,2048,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,2560,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,2560,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,2560,0.013439999686347114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,2048,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,2048,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,1536,0.003942222230964237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,2048,0.01310222182008955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,5120,0.008617777791288164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,1536,0.006729777902364731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,5120,0.020011555817392137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,1536,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,768,0.0030417777597904205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,1536,0.011702222128709158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,1024,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,1024,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,1024,0.010416000253624385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,512,0.0027555556346972785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,768,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,768,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,768,0.0103493332862854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,256,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,512,0.005663110978073544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,512,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,128,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,512,0.010059555371602377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,256,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,256,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,64,0.002116444417172008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,256,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,32,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,128,0.005632888939645555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,128,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,65536,0.04714577727847629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,128,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,64,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,32,0.005160000175237656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2560,4096,0.0074648890230390764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,65536,0.055083556307686694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,65536,0.055777774916754834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,16384,0.015339555011855232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,65536,0.13502932919396296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,51200,0.04494755466779073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,51200,0.04378666811519199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,12288,0.013777777552604675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,16384,0.018248000078731112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,16384,0.017398221625222098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,51200,0.10734577973683675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,16384,0.04098755452367995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,12288,0.014304000470373364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,12288,0.014489778214030795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,12288,0.03322044346067641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,8192,0.011393778026103973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,10240,0.012533333566453723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,10240,0.01277511070171992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,7168,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,10240,0.029291556941138372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,8192,0.010838222172525195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,8192,0.011361777782440186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,8192,0.025557332568698462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,7168,0.010010666317409938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2560,1024,0.003216888962520493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,51200,0.038262221548292376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,7168,0.02300888962215847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,5120,0.008922666311264038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,6144,0.009639999932712978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,6144,0.021020443903075323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,5120,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,5120,0.00865955568022198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,5120,0.018580444984965853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,10240,0.012327111429638334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,3584,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,4096,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,4096,0.0075333333677715724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,4096,0.01651733285850949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,3072,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,3584,0.007958222594526079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,3584,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,3584,0.015463110473420886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,2560,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,6144,0.009497777455382878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,3072,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,7168,0.010140444669458602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,3072,0.014924445086055331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,2048,0.004519999855094486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,2560,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,2560,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,1536,0.0038604442444112566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,6144,0.009123555488056606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,2560,0.014319999350441827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,2048,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,1024,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,2048,0.01277511070171992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,4096,0.007435555259386699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,1536,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,1536,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,1536,0.011764444410800934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,1024,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,1024,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,1024,0.010614222122563256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,768,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,512,0.002487999904486868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,768,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,768,0.01036444471942054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,256,0.0022071111533376905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,512,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,3072,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,512,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,128,0.0021395554973019496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,256,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,256,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,256,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,128,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,2048,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,128,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,32,0.001971555459830496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,65536,0.03688800003793504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,128,0.00721688899728987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,768,0.0028799999919202593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,64,0.004958222309748332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,2048,32,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,51200,0.029458665185504492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,65536,0.054921779367658824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,65536,0.05302400059170193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,16384,0.01349955548842748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,65536,0.13448354932996961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,51200,0.04439111219512092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,512,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,12288,0.012326222327020435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,51200,0.10702311330371433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,16384,0.01682844426896837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,16384,0.01770755483044518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,10240,0.010664000279373594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,16384,0.040748443868425153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,12288,0.013977777626779346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,12288,0.032928887340757586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,10240,0.012192000117566852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,2048,64,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,10240,0.0120328888297081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,10240,0.029093331760830347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,8192,0.0107342220014996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,7168,0.009087111386987898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,8192,0.010420444111029306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,8192,0.024907555845048692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,6144,0.008289777570300633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,7168,0.010081777969996134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,7168,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,7168,0.022656889425383672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,51200,0.04213866591453552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,5120,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,6144,0.009662222531106737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,6144,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,4096,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,6144,0.020976000361972388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,5120,0.008960000342792934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,5120,0.008120888637171851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,5120,0.018545778261290658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,12288,0.013411555853154926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,4096,0.00831911133395301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,4096,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,3072,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,4096,0.017048888736300997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,3584,0.00794400026400884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,8192,0.009524444739023844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,3584,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,3072,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,3072,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,3584,0.015455999308162265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,2048,0.004299555387761858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,3072,0.015138667490747241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,2560,0.007251555720965068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,1536,0.0038364442686239877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,2560,0.013426666458447775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,2048,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,2048,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,1024,0.003149333306484752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,2048,0.012443555725945367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,1536,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,1536,0.0058808889653947614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,768,0.0028151110228565005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,1536,0.011704000333944956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,1024,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,1024,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,512,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,1024,0.010408889088365767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,768,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,768,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,256,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,768,0.010091555615266165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,512,0.005761777775155173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,512,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,128,0.0021271111650599376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,2560,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,512,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,256,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,256,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,64,0.0018791111393107309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,128,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,256,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,128,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,2560,0.006339555399285422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,65536,0.030435555511050757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,64,0.0052044445441828836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,128,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,51200,0.023873777853118047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1536,32,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,65536,0.05446044603983561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,65536,0.05050933361053467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,16384,0.011391110718250275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,51200,0.044128888183169894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,65536,0.13389689392513698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,3584,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,51200,0.10649689038594563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,12288,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,16384,0.016068443655967712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,16384,0.040884445110956825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,12288,0.013463111387358772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,12288,0.013000888956917657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,12288,0.03239466746648153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1536,32,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,10240,0.011431111229790581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,8192,0.008069333102968004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,10240,0.012073777616024017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,10240,0.02850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,7168,0.007334222396214803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,8192,0.010755555497275459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,8192,0.010057777994208867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,8192,0.024400000770886738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,6144,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,51200,0.04033777780003018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,7168,0.009480888644854227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,7168,0.022333333889643352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,5120,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,6144,0.009343999955389235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,16384,0.01614488826857673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,6144,0.008642666869693333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,4096,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,6144,0.020765332712067496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,5120,0.00868800034125646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,5120,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,10240,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,3584,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,5120,0.018251554833518136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,4096,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,4096,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,3072,0.005534222142563925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,4096,0.017172444197866652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,3584,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,2560,0.004859555512666702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,3584,0.015276443627145557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,3072,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,3072,0.0074444446298811175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,2048,0.0042071110672420925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,3072,0.01443911095460256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,2560,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,2560,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,1536,0.003556444413132138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,2560,0.01314399970902337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,7168,0.01001422190003925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,2048,0.006986666884687211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,2048,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,1024,0.003175111073586676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,2048,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,768,0.0028080000645584534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,1536,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,1024,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,1536,0.012082666986518435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,1024,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,512,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,768,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,1024,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,256,0.002151111140847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,768,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,512,0.0058577776783042485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,768,0.010330666270520952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,128,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,512,0.0052826665341854095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,512,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,256,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,3584,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,256,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,64,0.0018044445249769424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,1024,32,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,256,0.008973333570692275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,65536,0.023893333143658105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,128,0.00563466673096021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,128,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,51200,0.01942044496536255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,128,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,64,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,1024,32,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,65536,0.05411110983954536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,16384,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,65536,0.04959644542800056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,51200,0.04351377818319532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,51200,0.040018667777379356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,65536,0.1335715585284763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,16384,0.015863999724388123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,1536,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,51200,0.10628266466988458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,10240,0.007355555891990662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,16384,0.040425777435302734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,12288,0.013332444760534497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,12288,0.012780444489585029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,8192,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,10240,0.012059555285506778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,12288,0.03196177879969279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,10240,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,7168,0.007168889045715332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,10240,0.0288080010149214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,8192,0.01070844464831882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,8192,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,6144,0.006758222149478064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,8192,0.02401688860522376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,7168,0.010062221851613786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,7168,0.009319110876984065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,5120,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,7168,0.02274666726589203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,6144,0.008647110727098253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,6144,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,4096,0.006226666685607698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,6144,0.020263999700546265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,5120,0.008661333057615492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,5120,0.008015111088752747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,12288,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,3584,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,5120,0.01887999971707662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,16384,0.01520088811715444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,3072,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,3584,0.007848000360859765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,4096,0.016197333733240765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,3584,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,2560,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,3584,0.015167999598715039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,3072,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,3072,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,2048,0.004348444441954295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,3072,0.01516711049609714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,2560,0.0070168889231152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,2560,0.0064488889442549805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,1536,0.0036071112586392295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,2560,0.013448889056841532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,2048,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,2048,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,1024,0.0031795555518733132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,2048,0.013072888884279462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,1536,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,1536,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,768,0.00286666676402092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,1024,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,1024,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,1024,0.011077333655622272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,4096,0.0075804442167282104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,768,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,512,0.002493333278430833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,768,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,256,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,768,0.010052444206343757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,512,0.0057946667075157166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,128,0.002016000035736296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,512,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,256,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,512,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,256,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,64,0.0018213333355055915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,768,32,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,256,0.00904444439543618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,128,0.005399999933110342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,65536,0.01772444446881612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,128,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,128,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,64,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,51200,0.014516444669829475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,768,32,0.004948444250557157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,65536,0.05352444450060526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,16384,0.00793422261873881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,1536,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,65536,0.04861155483457777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,51200,0.04087377919091119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,65536,0.13319999641842312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,12288,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,16384,0.015791111522250705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,51200,0.10659555594126384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,16384,0.01510933372709486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,10240,0.006680000159475539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,16384,0.03972355524698893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,12288,0.012425777812798819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,8192,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,12288,0.03253600001335144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,10240,0.011931555966536203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,10240,0.011425777441925473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,7168,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,10240,0.02812533246146308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,8192,0.010724444356229572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,8192,0.009914666414260864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,6144,0.005983110931184556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,8192,0.02460622290770213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,7168,0.010031999813185798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,7168,0.009593778186374241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,7168,0.022630222969584998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,5120,0.006565333240562015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,6144,0.008646222452322641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,6144,0.009376889301670922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,6144,0.020085333122147452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,4096,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,5120,0.008743999732865227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,51200,0.03933777742915683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,5120,0.008275555239783393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,3584,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,5120,0.018220444520314533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,4096,0.008271111382378472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,12288,0.01311822235584259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,4096,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,3072,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,4096,0.01717333330048455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,3584,0.007727111379305522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,2560,0.005052444421582752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,3584,0.015198222464985318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,3072,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,3072,0.01440177857875824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,2048,0.004851555658711327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,2560,0.007051555646790399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,2560,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,2048,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,2048,0.006676444576846228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,2048,0.012464000119103326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,1536,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,1024,0.003175111073586676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,4096,0.00793599999613232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,1536,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,3584,0.007078222102589077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,768,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,1024,0.005993777679072485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,1024,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,1024,0.010416888528399997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,512,0.0024968888610601425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,768,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,2560,0.0064106666379504735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,768,0.010052444206343757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,256,0.0021679999513758556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,512,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,512,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,1536,0.003682666768630346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,512,0.0101742222905159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,128,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,256,0.00527555536892679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,256,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,64,0.0018231111268202465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,256,0.009017777939637503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,128,0.005535999933878581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,512,32,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,128,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,1536,0.011376889215575324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,64,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,32,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,51200,0.011254222028785281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,65536,0.0472924444410536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,65536,0.04653955499331156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,65536,0.13285333580440944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,16384,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,51200,0.03763644562827216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,51200,0.03682933251063029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,51200,0.10562222533755833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,16384,0.01569688816865285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,12288,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,16384,0.014906666345066495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,16384,0.03939466675122579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,10240,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,12288,0.013085333009560903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,12288,0.012455999851226807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,12288,0.03225244416130914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,65536,0.012496888637542725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,8192,0.005632888939645555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,10240,0.011750222080283694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,128,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,10240,0.011110222174061669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,7168,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,10240,0.027834667099846735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,8192,0.010468444062603844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,6144,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,8192,0.009941333697901832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,3072,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,7168,0.00905600024594201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,5120,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,7168,0.02199288871553209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,6144,0.009367999931176504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,6144,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,4096,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,6144,0.01995111174053616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,5120,0.008801778157552084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,5120,0.017955554856194388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,3584,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,4096,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,4096,0.016138666205936007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,3584,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,3584,0.00776800016562144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,3072,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,3584,0.015152000718646579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,3072,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,8192,0.0244151105483373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,2560,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,3072,0.014508444401952954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,3072,0.007382222347789341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,512,768,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,2560,0.007074666519959767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,2560,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,5120,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,2560,0.013423999978436364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,1536,0.0035582222044467926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,4096,0.00812266684240765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,2048,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,2048,0.012759110993809171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,1536,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,1024,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,1536,0.005743999862008625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,1536,0.012096888489193387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,768,0.0027057776848475137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,1024,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,1024,0.005383111122581694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,1024,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,512,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,768,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,768,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,768,0.010175111393133799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,256,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,512,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,512,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,2048,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,128,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,512,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,256,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,2048,0.006696888970004187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,256,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,64,0.0018079999006456798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,128,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,128,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,256,32,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,65536,0.009400889277458191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,128,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,64,0.005166222237878376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,256,32,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,65536,0.04618577824698555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,51200,0.008060444560315874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,65536,0.04513066675927904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,16384,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,65536,0.13226755460103354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,51200,0.036665777365366616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,12288,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,16384,0.01550222271018558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,51200,0.10513866609997219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,16384,0.03938044442070855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,10240,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,12288,0.013139555851618448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,12288,0.012701333396964602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,256,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,12288,0.0317831105656094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,8192,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,10240,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,10240,0.012076444096035428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,7168,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,10240,0.02773688899146186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,8192,0.010649777948856354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,8192,0.00978311151266098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,6144,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,7168,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,7168,0.02197422252760993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,51200,0.0373599992858039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,6144,0.008625778059164682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,5120,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,6144,0.009311111436949836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,6144,0.020281778441535104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,4096,0.004559111264016894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,5120,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,5120,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,5120,0.0188719994492001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,3584,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,4096,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,4096,0.00793511089351442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,4096,0.016135111451148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,3584,0.007721777591440413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,3072,0.004592888885074192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,3584,0.007163555257850223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,3584,0.01514400045077006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,2560,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,3072,0.007506666911972894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,3072,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,8192,0.024358222881952923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,7168,0.00963644435008367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,3072,0.01442311041884952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,2560,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,2560,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,1536,0.0035457776652442086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,2048,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,2048,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,7168,0.010018666585286459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,1024,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,2048,0.012737777498033313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,16384,0.014781332678265043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,1536,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,1536,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,768,0.002612444468670421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,1536,0.012063110868136087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,1024,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,1024,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,512,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,1024,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,768,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,768,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,256,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,768,0.01038577738735411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,512,0.005015999906592899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,512,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,128,0.001867555495765474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,512,0.009584888815879822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,256,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,256,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,64,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,256,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,128,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,32,0.0017813332378864288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,65536,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,128,0.006362666686375936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,2560,0.013824889229403602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,51200,0.007180444068378872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,64,0.004599110947714912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,128,32,0.004874666531880696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,12288,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,65536,0.045427554183536105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,10240,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,51200,0.036328001154793635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,16384,0.0147724449634552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,8192,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,7168,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,12288,0.012378666963842181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,6144,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,10240,0.011361777782440186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,5120,0.004558222161398994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,8192,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,4096,0.004469333423508538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,3584,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,6144,0.0086666668454806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,3072,0.004521777646409141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,128,2048,0.004679999831649992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,2560,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,4096,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,128,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,2048,0.0042035554846127825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,3584,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,1536,0.003547555456558863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,1024,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,3072,0.006768888897365994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,768,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,16384,0.005886222339338726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,2560,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,512,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,2048,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,256,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,1536,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,128,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,1024,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,64,0.001608000033431583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,64,32,0.0015475555426544612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,768,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,65536,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,256,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,128,0.004592000196377437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,51200,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,64,0.004580444345871608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,16384,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,32,0.004519111166397731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,12288,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,10240,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,65536,0.04490400022930569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,7168,0.009665778113736046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,8192,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,51200,0.03600177831119961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,16384,0.014780445231331719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,12288,0.012392000191741519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,6144,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,5120,0.004580444345871608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,5120,0.008088000118732452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,10240,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,4096,0.004226666771703296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,3584,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,8192,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,3072,0.004585777719815572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,7168,0.009032888544930352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,2560,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,6144,0.008356444537639618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,2048,0.004243555582231945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,5120,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,1536,0.0033271111961868075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,3584,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,1024,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,3072,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,768,0.0025680000997251934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,2560,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,2048,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,512,0.0021057778762446507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,1536,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,64,512,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,256,0.0018275556051068837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,1024,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,128,0.001791999985774358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,64,0.0017831110292010838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,768,0.005502222312821283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,32,0.001806222316291597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,512,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,256,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,128,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,64,32,7168,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,64,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,32,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,16384,0.36608976787990993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,16384,0.16630310482449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,64,32,4096,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,51200,0.4953395525614421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,16384,0.13331911298963758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,12288,0.2741280131869846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,12288,0.12406844562954372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,10240,0.22878487904866537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,12288,0.1163182258605957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,51200,1.1332239574856229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,10240,0.10514666636784871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,8192,0.18646844228108725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,10240,0.08620088630252415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,7168,0.162538660897149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,8192,0.0709279974301656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,6144,0.14132711622450086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,7168,0.06291821930143568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,8192,0.0882915523317125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,6144,0.054766221178902515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,5120,0.12119910452100967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,5120,0.04618933465745714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,5120,0.055679109361436635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,4096,0.09589778052435981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,3584,0.08538133568233913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,6144,0.0658551123407152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,4096,0.03912711143493652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,3072,0.07189866569307116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,4096,0.04656889041264852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,3584,0.04126844472355313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,2560,0.06087466743257311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,3584,0.035980443159739174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,3072,0.03627289003796048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,2048,0.05038933290375603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,1536,0.03881066706445482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,2560,0.027124444643656414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,2560,0.03215466605292426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,2048,0.0229404436217414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,7168,0.07518666982650757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,51200,0.4218497806125217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,2048,0.027348443865776062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,1536,0.019124444988038804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,1536,0.022023111581802368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,1024,0.015122666954994202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,512,0.014676445060306124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,1024,0.016501333978441026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,256,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,768,0.01368977791733212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,512,0.012423111332787408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,128,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,512,0.011072888970375061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,256,0.010695111420419482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,64,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,256,0.009071110851234859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,32,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,128,0.010695999695195092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,128,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,64,0.009584888815879822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,3072,0.03131999903255039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,65536,32,0.012480888929631976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,1024,0.027507555153634813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,65536,768,0.02128711177243127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,65536,768,0.013471999930010902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,51200,0.39676621225145126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,65536,0.5223591062757704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,16384,0.27696177694532603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,65536,1.0886347028944228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,51200,0.8503537707858615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,12288,0.208116438653734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,16384,0.13364622328016493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,16384,0.10688889026641846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,12288,0.09990844461652969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,8192,0.14179910553826228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,10240,0.07003821929295857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,10240,0.17306221856011283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,10240,0.08522133032480876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,65536,0.4303902255164252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,12288,0.08939555618498062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,8192,0.07192799780103895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,6144,0.105968885951572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,7168,0.06138222085105049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,6144,0.05379911263783773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,5120,0.08957688675986396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,6144,0.04410577813784281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,4096,0.07269244723849826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,5120,0.045737776491377086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,5120,0.03828977876239353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,4096,0.037802666425704956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,3584,0.0641768905851576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,4096,0.032487998406092324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,3072,0.056481778621673584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,3584,0.029228445556428697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,3584,0.034204443295796715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,2560,0.04703911145528158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,3072,0.025921778546439275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,3072,0.030232889784706965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,7168,0.1242106623119778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,2560,0.026591110560629103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,8192,0.05706044700410631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,2560,0.022610666023360357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,1536,0.030121776792738173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,2048,0.022651554809676275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,1024,0.02219466699494256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,1536,0.016392888294325936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,1536,0.018231110440360177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,768,0.01607022186120351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,1024,0.013432888521088494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,1024,0.013797333670987023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,512,0.011416888899273343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,768,0.01239822225438224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,768,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,256,0.00741511086622874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,512,0.010749333434634738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,512,0.010279999838935005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,128,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,51200,0.3320355680253771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,256,0.009509333305888707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,2048,0.03838133480813768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,64,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,256,0.007867555651399825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,51200,32,0.0042151109211974675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,7168,0.051408890220854014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,128,0.009710222482681274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,128,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,64,0.008639111287064022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,51200,2048,0.019494222270117868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,51200,32,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,65536,0.3461502128177219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,16384,0.08271466361151801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,51200,0.27270666758219403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,65536,0.18165689044528535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,12288,0.06341599755816989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,51200,0.1344711118274265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,65536,0.1486711104710897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,16384,0.04853689008288913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,16384,0.0421653323703342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,51200,0.116413328382704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,12288,0.037343111303117536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,8192,0.04418399930000305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,7168,0.03794488973087735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,10240,0.02989333205752903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,6144,0.03320533368322585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,8192,0.02374222212367588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,10240,0.0322053333123525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,8192,0.028110222684012517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,7168,0.021828444467650518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,5120,0.02880000074704488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,7168,0.024365334047211543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,4096,0.024104000793562994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,6144,0.01939288940694597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,3584,0.020607110526826646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,5120,0.01749955614407857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,4096,0.015368888775507608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,5120,0.018945778409639995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,3072,0.01850666602452596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,4096,0.015838222371207345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,3584,0.014471999473041959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,2560,0.015453333655993143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,3584,0.014550222290886773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,2048,0.012703110774358114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,3072,0.013193777865833707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,2560,0.01239733315176434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,1536,0.010208889014191097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,2048,0.010976889067225985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,10240,0.05255377623769972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,1024,0.007909333540333642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,2048,0.011344888971911537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,1536,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,768,0.006342222293217977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,1024,0.007824888659848107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,1024,0.007848888635635376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,768,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,512,0.004831999954250124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,768,0.006672888994216919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,256,0.0031946665710873077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,6144,0.02202400068442027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,512,0.0069884442620807225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,128,0.00282222218811512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,256,0.006593777901596493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,256,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,64,0.0025128889828920364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,3072,0.013086222112178802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,16384,32,0.002597333242495855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,128,0.00693866651919153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,16384,2560,0.01221866657336553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,128,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,64,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,32,0.006562666760550604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,12288,0.03727466530270047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,65536,0.15151377518971762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,51200,0.2117155525419447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,16384,0.07532088624106513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,65536,0.12954933113522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,1536,0.009768000079525841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,12288,0.05189777745140923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,51200,0.10564533207151626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,16384,0.03751466671625773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,16384,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,16384,0.04234755701488919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,10240,0.04284088810284933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,12288,0.03288088904486762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,12288,0.03139822350607978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,10240,0.025853332546022203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,7168,0.031160887744691636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,8192,0.0214720004134708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,8192,0.02477600011560652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,6144,0.0270106659995185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,7168,0.019559999306996662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,7168,0.021529777182473078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,5120,0.023129777775870428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,6144,0.01772711177666982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,6144,0.019278221660190158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,4096,0.01994222237004174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,51200,0.11603821648491754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,5120,0.01625955601533254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,5120,0.016537777251667447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,4096,0.014186667071448432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,4096,0.01443377799457974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,3072,0.015084443820847405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,3584,0.013408888545301227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,2560,0.012520000338554382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,3584,0.01332800007528729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,8192,0.03601155678431193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,2048,0.01069777790043089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,3072,0.01238933371173011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,2560,0.011547555526097616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,1536,0.008622222476535374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,10240,0.028466665082507666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,2560,0.01129422254032559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,1024,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,2048,0.009663110805882348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,2048,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,1536,0.008352888955010308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,65536,0.2697288990020752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,768,0.0057546666098965546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,512,0.00416355538699362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,1536,0.009022222624884712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,1024,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,256,0.0029884444342719186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,768,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,512,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,768,0.006303999986913469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,128,0.002510222295920054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,512,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,256,0.0064080001579390625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,64,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,3584,0.01702311138312022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,12288,32,0.002691555561290847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,128,0.0063564446237352155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,256,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,128,0.005435555759403441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,64,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,12288,32,0.0064488889442549805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,65536,0.21810489230685762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,3072,0.012273777690198688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,51200,0.17217332786983916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,16384,0.061984890037112765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,65536,0.14580178260803223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,65536,0.12192532751295303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,12288,0.04142133394877116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,51200,0.11063377724753486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,10240,0.034091555409961276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,12288,0.02989333205752903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,16384,0.03886311252911886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,16384,0.04039466712209914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,51200,0.09916177723142837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,8192,0.02888800038231744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,7168,0.024887111451890733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,12288,1024,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,10240,0.02476355599032508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,6144,0.021613332960340712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,8192,0.020733333296246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,7168,0.018964444597562153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,5120,0.01891911029815674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,8192,0.023248000277413264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,7168,0.020378667447302077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,6144,0.01718933383623759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,5120,0.015454222758611044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,6144,0.01813066667980618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,3584,0.013744889034165276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,4096,0.013782222237851886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,5120,0.01606399979856279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,3072,0.012056888805495368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,4096,0.01408266690042284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,2560,0.01033866653839747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,3072,0.011972444752852121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,3584,0.0131217779384719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,3072,0.012087110843923358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,2048,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,2560,0.01070844464831882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,2560,0.010951110886202918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,1536,0.007244444555706448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,2048,0.009341333475377824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,2048,0.010404444403118556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,12288,0.0311173333062066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,1024,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,1536,0.00831288927131229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,768,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,1536,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,10240,0.027089777919981215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,1024,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,512,0.003522666792074839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,4096,0.01607288916905721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,256,0.0031475555151700974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,768,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,768,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,512,0.006766222417354584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,128,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,512,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,64,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,256,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,128,0.006567111031876669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,128,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,64,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,32,0.006330666856633292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,65536,0.09416355689366658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,65536,0.0696657763587104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,51200,0.14104444450802273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,10240,1024,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,16384,0.048230220874150596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,51200,0.07208977805243598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,51200,0.056573331356048584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,12288,0.035783999496036105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,256,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,10240,0.029855112234751385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,16384,0.02843822373284234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,16384,0.02258666687541538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,10240,32,0.0025235555238193935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,8192,0.025384000606007043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,12288,0.019137778215938143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,10240,3584,0.013002666334311167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,7168,0.021894221504529316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,10240,0.016368000043763053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,6144,0.019162666466501024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,10240,0.019328888919618394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,7168,0.013481777575280933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,8192,0.01612444387541877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,5120,0.016335111525323655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,7168,0.014764444695578681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,6144,0.012438221938080259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,4096,0.014091556270917257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,6144,0.013545778062608508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,5120,0.012423999607563019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,3584,0.012057777908113269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,5120,0.012068444656001197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,4096,0.010682666467295753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,3072,0.010715555813577441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,4096,0.010831111007266574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,3584,0.009597332941161262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,2560,0.009510222408506606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,3072,0.008961777720186446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,3072,0.00943022221326828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,2048,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,2560,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,1536,0.006743999818960826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,2048,0.0075937774446275495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,2048,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,12288,0.022032888399230108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,1024,0.004976888911591636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,1536,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,768,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,8192,0.01444355481200748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,1024,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,65536,0.178110228644477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,512,0.003162666741344664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,1024,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,768,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,512,0.006072888771692912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,768,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,128,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,512,0.0052844443255000645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,256,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,256,0.00508355556262864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,64,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,32,0.002326222136616707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,128,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,128,0.005162666655249066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,3584,0.01036266651418474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,2560,0.008284444610277811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,64,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,65536,0.16063111358218723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,8192,32,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,51200,0.12714400556352404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,16384,0.045612444480260216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,65536,0.08769777748319839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,65536,0.0670293304655287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,8192,1536,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,12288,0.0358222226301829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,51200,0.06569333208931817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,51200,0.054377777708901294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,16384,0.021779555413458083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,10240,0.027407111393080816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,16384,0.026032888227038916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,12288,0.017885333961910672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,8192,0.023919110496838886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,7168,0.02050844497150845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,8192,256,0.002549333290921317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,12288,0.02029600077205234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,10240,0.015793777174419828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,6144,0.01776266594727834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,10240,0.017871111631393433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,5120,0.015206222732861837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,7168,0.013051555388503604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,6144,0.01201866649919086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,6144,0.01278133359220293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,4096,0.013088000317414602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,5120,0.011101333631409539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,3584,0.011711999773979187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,5120,0.011381333072980246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,4096,0.009737778041097853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,3072,0.01034311122364468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,3584,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,4096,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,2560,0.008960888617568547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,3584,0.009903111391597325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,3072,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,2048,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,3072,0.009136888715955947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,2560,0.007700444095664554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,1536,0.006539555473460092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,2560,0.008468444148699442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,2048,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,1024,0.0047866664826869965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,2048,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,1536,0.006761777732107375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,1536,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,768,0.003567999849716822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,8192,0.014169777433077494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,1024,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,512,0.0031928889867332247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,8192,0.015522667103343539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,1024,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,256,0.0025395556456512874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,768,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,128,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,512,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,768,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,64,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,7168,32,0.002521777732504739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,128,0.0058613332609335584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,256,0.005111111121045219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,128,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,65536,0.13965422577328152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,64,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,32,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,51200,0.10991199811299641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,65536,0.07806044154696994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,65536,0.06439822249942355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,16384,0.03820977939499749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,51200,0.05247911148601108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,16384,0.02091911103990343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,16384,0.024274667104085285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,10240,0.026437333888477747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,12288,0.018566222654448617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,10240,0.015311110350820752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,7168,0.013757333159446716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,8192,0.02344266739156511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,7168,512,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,7168,0.020562666985723708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,8192,0.013567999833159976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,8192,0.0147724449634552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,7168,0.01274222218328052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,7168,0.013081777426931592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,6144,0.011735999749766456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,6144,0.012135999898115793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,5120,0.01418133411142561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,5120,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,4096,0.012376888758606382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,12288,0.01722666621208191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,5120,0.010795555180973478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,3584,0.010674667027261523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,4096,0.009193778038024902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,10240,0.016182222300105624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,7168,256,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,4096,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,3584,0.009489778015348647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,2560,0.00832799987660514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,3072,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,3072,0.008741333252853816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,6144,0.018186666899257235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,2560,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,2048,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,2048,0.0064071110553211635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,1536,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,1536,0.006873777757088344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,1536,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,1024,0.003900444342030419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,3072,0.009655111365848118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,1024,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,3584,0.008341333104504479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,768,0.0035386664999855887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,1024,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,512,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,768,0.006123555617200003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,2048,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,768,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,256,0.002475555572244856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,512,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,128,0.0021991110924217436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,512,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,256,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,51200,0.05936621957355075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,64,0.002151111140847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,32,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,256,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,128,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,128,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,65536,0.11722311708662246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,64,0.005414222263627582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,6144,32,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,51200,0.09293066793017918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,65536,0.07130399677488539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,65536,0.06215200159284803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,16384,0.03373600045839945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,12288,0.02666666607062022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,51200,0.05387377738952637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,16384,0.020281778441535104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,16384,0.02204622162712945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,10240,0.022079111801253423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,12288,0.01661155621210734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,8192,0.018360889620251127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,12288,0.01682933337158627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,10240,0.014875556031862894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,10240,0.0151555554734336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,7168,0.016538666354285348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,8192,0.013218666944238873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,6144,12288,0.03086844417783949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,6144,0.01516266663869222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,8192,0.013447999954223633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,5120,0.013227555486891003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,7168,0.01238844460911221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,6144,0.011239999698268043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,4096,0.011375111010339526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,5120,0.009728000395827824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,5120,0.01035822182893753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,3584,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,4096,0.00835111074977451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,4096,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,3072,0.009321778184837764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,51200,0.050159109963311084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,3584,0.008072888685597314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,2560,0.007927999728255803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,3584,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,3072,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,2048,0.005979555762476391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,2560,0.007337777978844113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,3072,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,1536,0.004850666556093427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,6144,2560,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,2560,0.0069368887278768755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,2048,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,1024,0.0036915557252036203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,768,0.0035182221068276297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,2048,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,1536,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,1024,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,1536,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,7168,0.01237155579858356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,768,0.005990222096443176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,1024,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,256,0.002456888970401552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,768,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,128,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,512,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,64,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,256,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,6144,0.011328000161382886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,32,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,5120,128,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,256,0.004987555659479565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,128,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,65536,0.09513244363996719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,64,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,51200,0.07257244322035047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,32,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,16384,0.027775999572541978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,65536,0.06020800272623698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,65536,0.06403644217385186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,12288,0.02139644490347968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,51200,0.04825244347254435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,51200,0.04951999915970696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,16384,0.01958311100800832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,10240,0.01866133345497979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,8192,0.01650311052799225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,12288,0.016117332710160148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,12288,0.015608888533380298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,10240,0.014498665928840637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,7168,0.01423911088042789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,10240,0.014139556222491793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,5120,512,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,8192,0.01276711126168569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,6144,0.011861333416567909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,8192,0.013059555656380124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,5120,0.010267555713653564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,6144,0.010573333336247338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,7168,0.01146399974822998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,5120,512,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,6144,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,5120,0.009478222164842818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,5120,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,4096,0.008289777570300633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,4096,0.008621333373917473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,3072,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,3584,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,3584,0.007827555967701806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,2560,0.0063528890411059065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,3072,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,3072,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,2048,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,2560,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,2560,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,1536,0.004271111140648524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,2048,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,2048,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,7168,0.011568889021873474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,1536,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,1536,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,768,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,4096,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,512,0.0025235555238193935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,1024,0.006224888894293044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,3584,0.007925333248244392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,1024,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,768,0.005990222096443176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,256,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,128,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,512,0.005882666756709416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,256,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,64,0.001869333287080129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,4096,128,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,32,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,128,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,65536,0.07697866360346477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,64,0.004955555415815777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,16384,0.019510222805870902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,32,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,51200,0.06193777587678698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,65536,0.05886489152908325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,65536,0.06191466914282905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,4096,1024,0.0032400000426504347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,51200,0.04725422130690681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,51200,0.047934220896826855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,12288,0.019882667395803664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,16384,0.019280888968043856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,16384,0.01864177816443973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,10240,0.017077333397335477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,12288,0.015816888875431485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,8192,0.014744000302420722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,12288,0.01514311134815216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,768,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,10240,0.014463999205165438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,7168,0.012674666941165924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,512,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,10240,0.013749333719412485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,8192,0.012373333175977072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,4096,256,0.004945777770545748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,8192,0.012605333493815528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,7168,0.011372444530328115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,7168,0.011258666714032492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,4096,0.008761777646011775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,5120,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,6144,0.010498666928874122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,3584,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,5120,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,4096,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,3072,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,4096,0.007962666451931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,3584,0.00793333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,2560,0.006214222146405115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,3584,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,3072,0.007702222300900354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,2048,0.005116444494989184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,2560,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,3072,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,1536,0.0041804446114434134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,2560,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,2048,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,1024,0.0032186667538351486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,2048,0.006324444380071428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,1536,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,768,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,1536,0.005998222364319696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,6144,0.011090666883521609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,512,0.002549333290921317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,5120,0.009748444788985783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,1024,0.005655999812814925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,768,0.005970666805903117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,256,0.0025031111306614345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,768,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,512,0.005877333382765452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,512,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,256,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,256,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,64,0.0019253332995706135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,128,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,32,0.0020684444655974707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,128,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,16384,0.024757333927684363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,65536,0.06853244702021281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,64,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3584,32,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,51200,0.05446399913893806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,65536,0.059337774912516274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,65536,0.057871997356414795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,16384,0.02092444399992625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,51200,0.046592000457975596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,16384,0.018930666976504855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,1024,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,12288,0.01750666730933719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,16384,0.017986666825082567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,10240,0.014765333798196582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3584,6144,0.009993777506881291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,12288,0.01583377851380242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,8192,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,12288,0.014818666709793938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3584,128,0.0021235555824306277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,7168,0.012073777616024017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,10240,0.013366222381591797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,8192,0.011618666350841522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,6144,0.010942222343550788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,8192,0.012078222301271228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,7168,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,5120,0.009506666825877296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,7168,0.011000889043013254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,6144,0.009679999616410997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,4096,0.008273777862389883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,6144,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,3584,0.007251555720965068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,5120,0.009183111290136972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,4096,0.008282666405042013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,3072,0.006790222393141852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,4096,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,3584,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,2560,0.005867555737495422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,3072,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,3584,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,2048,0.004676444249020683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,3072,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,51200,0.04619377851486206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,1536,0.004221333397759331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,2560,0.007419555551475949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,2048,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,1024,0.0032302221904198327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,1536,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,2048,0.006219555520349079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,768,0.003154666680428717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,1536,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,512,0.002817777709828483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,1024,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,768,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,1024,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,256,0.002552888873550627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,768,0.00553955551650789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,512,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,128,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,512,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,256,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,64,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,5120,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,3072,32,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,128,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,256,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,65536,0.058055109447903104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,128,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,51200,0.04720711045795017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,65536,0.05679821968078613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,32,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,65536,0.056327998638153076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,51200,0.0455022222465939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,51200,0.04436444573932224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,16384,0.018908444378111098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,10240,0.013605333036846586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,16384,0.017522666189405654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,3072,2560,0.006797333144479328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,12288,0.015032000011867948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,8192,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,3072,10240,0.013938667045699226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,12288,0.01443733274936676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,7168,0.011237333218256632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,8192,0.011083555718262991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,10240,0.0131217779384719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,6144,0.01020088874631458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,8192,0.011682666838169098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,7168,0.01033155537313885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,5120,0.009209777745935652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,7168,0.010656000011497073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,6144,0.009694221946928237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,4096,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,6144,0.009679999616410997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,5120,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,3584,0.007174222005738153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,4096,0.008316444853941599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,4096,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,3072,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,16384,0.01925155520439148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,3584,0.00811288919713762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,2560,0.005511999958091312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,3584,0.007268444531493717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,12288,0.015798222687509324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,2048,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,3072,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,2560,0.00666133314371109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,1536,0.003996444245179494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,2048,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,10240,0.013399110900031196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,2048,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,1024,0.003212444484233856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,1536,0.006728888799746831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,768,0.0029484445436133277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,1024,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,1536,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,512,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,1024,0.005538666827811136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,768,0.005971555494599872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,256,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,768,0.005482666608360078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,512,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,128,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,512,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,256,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,64,0.0021911110315057966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,256,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,128,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2560,32,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,128,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,3072,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,65536,0.049599998527103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2560,2560,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,64,0.004951111144489712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,32,0.005283555636803309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,16384,0.016193777322769165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,65536,0.05418044328689575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,51200,0.04486133323775398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,12288,0.013817778064144982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,51200,0.04259022076924642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,16384,0.018312000566058688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,16384,0.01719999975628323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,10240,0.01238844460911221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,12288,0.014088888963063559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,8192,0.011767999993430244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,12288,0.014115555418862237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,10240,0.013097777962684631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,7168,0.010235555469989777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2560,5120,0.008460444708665213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,10240,0.01275199982855055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,8192,0.011037333144081963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,6144,0.009670221971140968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,8192,0.011015999648306104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,5120,0.008757332960764566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,7168,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,7168,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,6144,0.009717333647939894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,6144,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,5120,0.0086986662613021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,3584,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,4096,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,5120,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,51200,0.03946400019857619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,3072,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,4096,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,3584,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,65536,0.05572711096869575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,2048,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,3584,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,3072,0.006875555548402999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,1536,0.0038604442444112566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,2560,0.006528000036875407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,2048,0.007096000015735626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,1024,0.003198222153716617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,2048,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,1536,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,768,0.0028106667515304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,1536,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,1024,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,512,0.002532444480392668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,1024,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,768,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,256,0.002206222257680363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,768,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,512,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,128,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,512,0.0053173332578606075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,256,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,64,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,256,0.005071111023426056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,128,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,32,0.0018737777653667664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,128,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,2560,0.004965333475006951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,65536,0.038266665405697294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,3072,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,64,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,2048,2560,0.007328000333574083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,2048,32,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,16384,0.01348000019788742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,65536,0.05509688787990146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,51200,0.04434222314092848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,51200,0.0410480002562205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,16384,0.01733688844574822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,16384,0.016759999924235873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,10240,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,12288,0.013415111435784234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,12288,0.013808888693650564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,8192,0.009491555392742157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,10240,0.012216000093354119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,2048,4096,0.007714666426181793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,10240,0.012178666889667511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,7168,0.009022222624884712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,8192,0.010808889236715106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,8192,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,6144,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,7168,0.009456000394291347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,5120,0.007876444194051955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,6144,0.009662222531106737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,51200,0.030408889055252075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,6144,0.008816888762844933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,4096,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,5120,0.008983111215962304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,65536,0.05196355448828804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,3584,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,5120,0.008057778080304464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,3072,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,4096,0.008126222425036961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,12288,0.012103999654452005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,2560,0.004861333303981357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,4096,0.007326222128338284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,3584,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,2048,0.004564444637960858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,3072,0.007592000067234039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,3584,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,2560,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,3072,0.006776000062624614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,1024,0.0031822222388452957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,2048,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,2560,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,768,0.002834666727317704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,2048,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,1024,0.006077333456940121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,1536,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,512,0.0024844445288181305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,768,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,1024,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,256,0.0021911110315057966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,768,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,512,0.005656888915432825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,7168,0.009995555712117089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,128,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,64,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,256,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,512,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,32,0.0018577778504954444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,256,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,65536,0.027600889404614765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,128,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,51200,0.023710222707854375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,64,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1536,32,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,65536,0.0544453329510159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,16384,0.010991111397743225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,65536,0.04934844374656677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,12288,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,51200,0.04388622111744351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,16384,0.01588711142539978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,51200,0.03958844476275974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1536,1536,0.00388355553150177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,10240,0.008655110994974772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,12288,0.013292444248994192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,8192,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,12288,0.012747555143303342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,1536,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,7168,0.007424888511498769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,8192,0.0107351111041175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,10240,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,6144,0.009669333696365356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,8192,0.010041777458455827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,7168,0.010103999740547603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,5120,0.008299555215570662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,7168,0.00960355583164427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,4096,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,6144,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1536,128,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,5120,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,3584,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,4096,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,4096,0.00758133331934611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,3584,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,3584,0.007234666910436418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,3072,0.007360888852013483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,2560,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,3072,0.006712889091836081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,16384,0.0162471118900511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,2048,0.004366222355100843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,2560,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,10240,0.012025777664449481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,1536,0.0038453332251972626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,2048,0.006851555572615729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,2048,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,1024,0.003167111012670729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,6144,0.009543110926946005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,1536,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,1536,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,5120,0.00868977771864997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,1024,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,512,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,768,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,1024,0.005764444255166584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,256,0.002157333410448498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,768,0.005356444252861871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,3072,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,512,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,128,0.0021537778278191886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,256,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,64,0.0018231111268202465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,1024,128,0.00564533347884814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,32,0.0019031111150979996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,128,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,65536,0.02337155573897892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,64,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,32,0.005055111315515307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,2560,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,51200,0.019262222780121695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,65536,0.05412888858053419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,65536,0.048454221751954824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,16384,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,51200,0.043806221750047475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,1024,768,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,12288,0.008253333469231924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,51200,0.03903466794225905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,16384,0.015835555063353646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,10240,0.007340444458855524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,16384,0.015056000815497505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,12288,0.013413333230548434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,8192,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,10240,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,7168,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,12288,0.012518222133318583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,10240,0.011295999917719098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,8192,0.010452444354693094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,6144,0.006215111249023014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,512,0.005294222384691238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,7168,0.010032888915803697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,8192,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,5120,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,4096,0.006734222173690796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,7168,0.009341333475377824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,6144,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,3584,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,5120,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,6144,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,3072,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,4096,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,5120,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,3584,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,2560,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,4096,0.007476444045702617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,3072,0.00758133331934611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,3584,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,1536,0.003730666720204883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,2560,0.007131555842028723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,3072,0.006789333290523953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,1024,0.003126222226354811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,2048,0.006911110960774952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,2560,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,768,0.0028435554769304064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,1536,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,2048,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,512,0.002536888958679305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,1536,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,768,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,1024,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,256,0.002161777681774563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,512,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,768,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,128,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,512,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,256,0.005325333525737126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,64,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,32,0.0018764444523387486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,256,0.004967111266321606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,65536,0.0173653331067827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,128,0.004862222406599256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,1024,256,0.005118222286303838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,64,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,51200,0.014157333307796054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,768,32,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,65536,0.05218133330345154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,16384,0.007888889147175683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,65536,0.04794044295946757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,768,2048,0.004228444563017951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,51200,0.039381331867641874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,12288,0.0070364442136552595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,16384,0.015840889679061044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,10240,0.0064151109092765385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,16384,0.01478488908873664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,12288,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,8192,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,12288,0.0124195557501581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,7168,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,10240,0.011142222417725457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,8192,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,6144,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,8192,0.009719111025333405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,7168,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,5120,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,7168,0.009511999785900116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,768,128,0.005538666827811136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,6144,0.009333333207501305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,4096,0.00656088896923595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,6144,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,5120,0.008961777720186446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,3584,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,5120,0.007968000239796108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,3072,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,4096,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,3584,0.007862221863534715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,2560,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,3584,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,3072,0.007377777662542131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,51200,0.03864444295565287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,2048,0.0041991112132867174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,3072,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,2560,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,1536,0.003705777641799715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,2560,0.006440889090299606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,2048,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,1024,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,1536,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,2048,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,10240,0.011759999725553723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,768,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,512,0.002502222235004107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,1536,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,1024,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,256,0.0021155555215146807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,768,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,1024,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,128,0.0019057778020699818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,512,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,768,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,256,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,512,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,128,0.005335999859703912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,256,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,65536,0.014151111245155334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,128,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,51200,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,64,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,512,4096,0.008007999923494127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,65536,0.04662044511901008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,16384,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,65536,0.046600000725852124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,51200,0.03738133443726434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,12288,0.0070293330483966405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,16384,0.015772443678643968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,51200,0.03606577714284261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,10240,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,16384,0.01477777792347802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,12288,0.013228444589508904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,8192,0.0068959999415609576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,10240,0.011888888974984487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,12288,0.01240711079703437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,7168,0.006559999866618051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,10240,0.011021333436171213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,6144,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,8192,0.010011555420027839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,7168,0.0099653336736891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,5120,0.006326222171386083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,7168,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,6144,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,64,0.001814222170246972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,4096,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,6144,0.008568888737095727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,512,32,0.001812444378932317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,5120,0.008712000317043727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,3584,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,3072,0.005304000029961268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,5120,0.00793422261873881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,4096,0.008053333395057255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,2560,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,4096,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,3072,0.007420444654093848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,3584,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,2048,0.0042204442951414315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,3072,0.006654222392373615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,2560,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,1536,0.003540444291300244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,2560,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,2048,0.006855999843941794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,1024,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,2048,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,1536,0.006384000182151794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,768,0.002891555635465516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,1536,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,1024,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,512,0.0024968888610601425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,1024,0.0053484443989064955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,768,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,256,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,768,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,512,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,128,0.0018702221827374564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,512,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,256,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,64,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,256,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,128,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,256,32,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,128,0.0047484445903036325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,3584,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,512,32,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,64,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,51200,0.010971555279360877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,256,32,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,65536,0.045900444189707436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,16384,0.005681777579916849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,65536,0.044158223602506846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,12288,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,51200,0.037364443143208824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,16384,0.015834665960735746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,51200,0.03584444522857666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,10240,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,16384,0.014615999327765571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,12288,0.01313333296113544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,8192,0.005535999933878581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,12288,0.012259555359681448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,256,8192,0.010399111443095738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,10240,0.0120666664507654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,7168,0.0053013331360287136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,6144,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,8192,0.010388444695207808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,10240,0.011387555963463254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,5120,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,7168,0.01014755583471722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,8192,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,4096,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,7168,0.009017777939637503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,6144,0.008607111043400234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,3584,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,5120,0.00815822184085846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,65536,0.012804444465372296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,4096,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,3584,0.00774133370982276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,2560,0.005061333378156026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,3072,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,3072,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,2048,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,2560,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,2560,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,1536,0.003530666646030214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,2048,0.00673333348499404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,2048,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,1024,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,1536,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,1536,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,768,0.0026168889469570587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,6144,0.009310222334331935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,1024,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,512,0.0024862223201327855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,1024,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,768,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,768,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,256,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,512,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,3072,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,128,0.001863111017478837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,512,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,256,0.005495111147562663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,3584,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,128,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,256,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,65536,0.011587555209795633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,128,0.004631999880075455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,51200,0.010039110978444418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,64,0.004593777573770947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,16384,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,128,32,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,12288,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,65536,0.04434400134616428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,10240,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,51200,0.03533333208825853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,8192,0.0052968888647026485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,16384,0.01479111115137736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,7168,0.00519822206762102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,12288,0.012436444560686747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,10240,0.011162666810883416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,6144,0.004948444250557157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,8192,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,5120,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,4096,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,7168,0.009319110876984065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,6144,0.008649778034951951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,64,0.0017751111752457088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,3584,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,128,32,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,3072,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,4096,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,3584,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,2048,0.00452888881166776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,3072,0.006670222100284364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,1536,0.003540444291300244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,2560,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,1024,0.0028844444702068963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,2048,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,768,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,1536,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,512,0.0022311111291249595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,1024,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,256,0.001968888979819086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,49,128,5120,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,128,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,768,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,64,0.001777777862217691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,4096,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,512,0.005006222261322869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,32,0.0017911110901170308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,256,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,65536,0.011548444628715515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,51200,0.00980444418059455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,128,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,5120,0.007744889292452071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,16384,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,64,0.004519111166397731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,64,32,0.004563555535342959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,12288,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,10240,0.005323555734422471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,65536,0.044180442889531456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,8192,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,7168,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,16384,0.014554666148291694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,12288,0.012067555553383298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,6144,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,10240,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,5120,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,4096,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,8192,0.009988444546858469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,7168,0.009020444419648912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,3584,0.004945777770545748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,3072,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,6144,0.008354666332403818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,2560,0.004594666676388847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,2048,0.004543999830881755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,4096,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,1536,0.0031840000301599503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,3072,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,1024,0.0028728888266616394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,2560,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,768,0.002467555511328909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,2048,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,512,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,1536,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,256,0.0018257778137922287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,1024,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,128,0.0018035556293196147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,64,2560,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,64,0.0018133332745896445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,512,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,49,32,32,0.0017928888814316855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,256,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,128,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,64,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,32,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,5120,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,3584,0.006990222467316522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,51200,0.4936658011542426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,16384,0.3643475638495551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,16384,0.1674506664276123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,768,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,51200,1.134129736158583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,49,32,51200,0.035599999957614474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,12288,0.27504444122314453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,12288,0.1241048839357164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,16384,0.13478755950927734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,12288,0.10815378030141194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,10240,0.2295306788550483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,51200,0.4014302094777425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,8192,0.18891999456617567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,10240,0.10484977563222249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,10240,0.08892177873187596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,8192,0.08919911252127753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,7168,0.16117421785990396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,8192,0.07111999723646376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,7168,0.07531555493672688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,7168,0.06326311164432101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,6144,0.13844888740115696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,16384,0.28470754623413086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,10240,0.17704533206091988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,12288,0.21283377541436088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,5120,0.1194231112798055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,6144,0.06605599986182319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,7168,0.12657154930962458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,8192,0.1422062185075548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,6144,0.06566488742828369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,4096,0.09639644622802734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,5120,0.055659552415212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,5120,0.04644977715280321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,4096,0.04658400019009908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,3584,0.08292088905970256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,4096,0.045888000064426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,5120,0.09508711099624634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,3584,0.041280001401901245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,6144,0.11102222071753608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,4096,0.0788959993256463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,3072,0.07248355282677545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,2560,0.0609111123614841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,3072,0.03066666589842902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,3072,0.036679999695883855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,3072,0.06310844421386719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,2560,0.03219377663400438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,2560,0.02683911058637831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,2048,0.048804445399178394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,2560,0.059839109579722084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,2048,0.023044443792766992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,1536,0.0398942232131958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,2048,0.02787289023399353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,2048,0.04969777663548788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,1024,0.028107557031843398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,1536,0.01884533299340142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,1536,0.022114665971861944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,768,0.020980444219377305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,1536,0.04219911164707608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,3584,0.04142933421664768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,1024,0.015122666954994202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,512,0.014575999644067554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,3584,0.07143377595477633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,1024,0.03366222315364414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,768,0.013526221944226159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,256,0.008355555435021719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,768,0.013802666631009845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,768,0.028840889533360798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,512,0.01204177737236023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,128,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,512,0.011125333607196808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,256,0.010505778094132742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,256,0.008982222113344405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,64,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,256,0.022127999199761286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,128,0.010770666930410596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,128,0.008005333443482717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,128,0.015770667129092746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,51200,0.8332871331108941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,64,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,32,0.012647111382749347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,65536,1024,0.01647644407219357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,65536,0.5226720174153646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,512,0.02439555525779724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,65536,1.1017867194281685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,51200,0.39647377861870664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,51200,0.8600017759535047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,65536,32,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,16384,0.27996354632907444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,16384,0.13388444317711726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,12288,0.21012443966335723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,16384,0.1067333353890313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,65536,0.42426665623982746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,10240,0.1758293310801188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,12288,0.10014488961961533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,12288,0.08906488948398167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,51200,0.3360782199435764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,10240,0.08521866798400879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,8192,0.14372889200846353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,10240,0.07825422286987305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,8192,0.07179111242294312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,8192,0.06526044342252943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,7168,0.12431022855970596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,16384,0.22442576620313856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,7168,0.061292443010542125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,10240,0.14395199881659612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,7168,0.05063377817471822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,12288,0.16899555259280732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,6144,0.10707822110917832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,5120,0.09152355459001328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,6144,0.04520444406403435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,7168,0.10505421956380208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,5120,0.039091553952958845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,5120,0.04565155506134033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,6144,0.09321955839792888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,4096,0.0738933351304796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,5120,0.08044444190131293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,4096,0.03787111242612203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,4096,0.03798755672242906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,4096,0.06772355238596599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,8192,0.11768000655704075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,3072,0.05654133028454251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,3584,0.03421955638461643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,3584,0.03404266635576884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,3584,0.06081866555743747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,6144,0.05396355523003472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,3072,0.03019022279315525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,2560,0.047592888275782265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,3072,0.026550221774313185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,3072,0.05537244346406725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,2048,0.03849511014090644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,2560,0.02630222174856398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,2560,0.060538669427235924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,51200,0.7112568749321831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,3584,0.06523911158243816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,2048,0.02304800020323859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,1536,0.031077331966824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,2048,0.019460444649060566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,2048,0.04443466663360596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,1024,0.02184088859293196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,1536,0.01847466660870446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,1536,0.01686044368478987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,768,0.016209777858522203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,1024,0.013815999858909182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,1024,0.013636444177892474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,1536,0.03647199935383267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,1024,0.02910755409134759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,512,0.011311111350854238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,65536,0.9019048478868273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,768,0.012318222059143914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,256,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,768,0.02500177754296197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,512,0.009924444059530893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,512,0.013125333521101209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,2560,0.025246222813924152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,512,0.021472889516088698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,128,0.00443822228246265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,768,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,64,0.003175111073586676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,256,0.009528000321653154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,51200,32,0.003504000190231535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,51200,128,0.01070666644308302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,256,0.019464888506465487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,128,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,51200,128,0.0140657772620519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,64,0.008667555948098501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,32,0.010842666857772403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,65536,0.3417368994818793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,65536,0.1742968824174669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,51200,0.27206844753689235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,16384,0.08275289005703397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,65536,0.1459475490781996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,51200,0.13464889261457655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,12288,0.06229244338141548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,16384,0.048455112510257296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,16384,0.042043556769688926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,51200,0.11951911449432373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,10240,0.0525511105855306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,51200,256,0.007608888877762689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,12288,0.037300444311565824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,12288,0.03487199876043532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,8192,0.04456088940302531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,16384,0.05697155661053128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,10240,0.032223112053341336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,10240,0.02808977829085456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,7168,0.03763822383350796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,8192,0.02404088940885332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,8192,0.02808977829085456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,8192,0.03530133432812161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,7168,0.021791110436121624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,6144,0.03287911083963182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,7168,0.024262222978803847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,51200,0.1411528852250841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,65536,0.1839279996024238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,5120,0.028229332632488672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,6144,0.022089777721299067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,6144,0.021898667017618816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,4096,0.02419288953145345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,6144,0.02836977773242527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,5120,0.0184879998366038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,5120,0.025049777494536504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,3584,0.020581333173645865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,4096,0.015196444259749519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,12288,0.044143110513687134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,4096,0.016229333149062265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,3072,0.018205333087179396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,4096,0.021746666895018682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,3584,0.01647733317481147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,10240,0.03901155458556281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,3584,0.014736889137162102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,2560,0.015218666858143277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,3584,0.020605333977275424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,3072,0.01330933305952284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,3072,0.013202666408485837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,2048,0.012558221817016602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,2560,0.011936888926559024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,3072,0.01853511068556044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,2560,0.017238222890430026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,7168,0.03078311019473606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,1536,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,2048,0.011051555474599203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,2048,0.010664888554149203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,1024,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,2048,0.016237333416938782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,2560,0.012105777859687805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,1536,0.009294222626421187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,768,0.006226666685607698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,1536,0.009816889133718278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,5120,0.01889866590499878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,1024,0.008051555189821456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,1536,0.014468444718254937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,512,0.004553777890072929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,1024,0.01297688898113039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,768,0.007393777370452881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,256,0.0028799999919202593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,768,0.011021333436171213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,512,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,512,0.006154666758245892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,128,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,512,0.01034755590889189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,256,0.006544888847404056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,256,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,64,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,256,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,128,0.005434666656785541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,16384,32,0.002467555511328909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,128,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,64,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,32,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,65536,0.25813423262702095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,65536,0.1503031121359931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,51200,0.20370488696628145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,1024,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,16384,768,0.006948444578382704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,65536,0.1333306630452474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,16384,0.07334133651521471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,51200,0.11634133921729194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,51200,0.10149600108464558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,16384,0.04222311245070564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,12288,0.05183910992410448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,16384,0.03922755519549052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,128,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,10240,0.04200622108247545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,16384,0.04765689041879442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,12288,0.030902223454581365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,12288,0.03277511066860623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,8192,0.03606933355331421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,12288,0.03790844480196635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,10240,0.025888888372315302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,65536,0.14673244953155518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,10240,0.02831288841035631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,7168,0.030986666679382324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,10240,0.033979554971059166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,8192,0.024303111765119765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,8192,0.021793777743975323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,51200,0.11648178100585938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,6144,0.02702666653527154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,8192,0.030957334571414526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,7168,0.02164711058139801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,7168,0.019762666689025033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,5120,0.02332177758216858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,6144,0.01796800063716041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,4096,0.019744889603720773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,7168,0.02735822233888838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,6144,0.019336000084877014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,5120,0.016526222229003906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,6144,0.025111999776628282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,5120,0.01640266676743825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,4096,0.01426755554146237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,4096,0.014490667316648694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,5120,0.0218231115076277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,3584,0.01424888935354021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,4096,0.019581332802772522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,3584,0.013395555317401886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,2560,0.012994666894276937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,3584,0.01847022275129954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,3072,0.01241422196229299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,3072,0.012172444827026792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,2048,0.010995555255148145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,3072,0.01741155485312144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,2560,0.012312889099121094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,2560,0.011352889239788055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,1536,0.008977777428097194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,2560,0.016210666961140104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,2048,0.010497777826256223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,1024,0.007323555648326874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,1536,0.008652444514963362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,2048,0.015384889311260648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,1536,0.00908622228437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,768,0.006053333481152852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,1536,0.014292443792025248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,1024,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,1024,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,512,0.004212444441186057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,1024,0.011686222420798408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,768,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,3584,0.016891555653678048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,768,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,256,0.0031742221779293488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,3072,0.014766222900814481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,512,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,128,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,512,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,512,0.009683555199040307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,256,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,256,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,256,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,64,0.0024906665914588505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,12288,32,0.002537777854336633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,128,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,128,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,2048,0.009775111244784461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,128,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,64,0.005994666781690385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,12288,32,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,65536,0.21600977579752603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,51200,0.17122488551669648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,65536,0.14188355869717068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,16384,0.06109778086344401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,65536,0.12160978052351211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,51200,0.11057777537239923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,768,0.01033688916100396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,51200,0.09824622339672512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,16384,0.040560000472598605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,16384,0.036208000448015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,12288,0.04169333312246534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,10240,0.034796443250444196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,16384,0.043988443083233304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,12288,0.029342220889197454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,12288,0.031154665682050917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,8192,0.029120888974931505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,12288,0.0358879996670617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,10240,0.02700533303949568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,10240,0.026331555512216356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,65536,0.1364497741063436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,10240,0.0322728885544671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,8192,0.022855111294322546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,8192,0.023061333431137934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,6144,0.022427555587556627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,8192,0.028404444456100464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,7168,0.020402666595247056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,5120,0.01882488860024346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,7168,0.026371555195914373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,6144,0.017362667454613578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,6144,0.01812266641192966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,4096,0.016093333562215168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,6144,0.02348711093266805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,5120,0.015606222881211175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,5120,0.01607644392384423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,3584,0.01408622165520986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,5120,0.021555556191338435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,4096,0.013963555296262106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,4096,0.019308444526460435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,3072,0.011975111232863532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,3584,0.013758222262064615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,3584,0.012829333543777466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,51200,0.1088906659020318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,3584,0.018213333355055917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,7168,0.024842666255103216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,3072,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,3072,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,2048,0.010693333215183683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,3072,0.01737955543729994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,2560,0.011412444214026133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,7168,0.01926399932967292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,2560,0.01075644459989336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,2560,0.0163128889269299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,2048,0.009299555586444007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,2048,0.010061333576838175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,1024,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,2048,0.01517688896920946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,1536,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,1536,0.014140443669425117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,1024,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,1024,0.007354666789372762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,4096,0.014855111638704935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,1024,0.011930666863918304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,512,0.004208000169859992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,768,0.0070462218589252895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,768,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,768,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,2560,0.01219377749496036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,512,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,256,0.0032355555643637977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,128,0.0028524444335036804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,512,0.009679999616410997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,256,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,256,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,64,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,1536,0.008638222184446123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,256,0.008970666262838576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,128,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,32,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,128,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,1536,0.008031999899281396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,65536,0.17424000634087455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,32,0.006230222268237009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,10240,64,0.0057813334796163775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,10240,768,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,65536,0.07118933068381415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,16384,0.048178666167789035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,51200,0.0719919999440511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,51200,0.05624533361858792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,16384,0.028424001402325098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,65536,0.13422311676873103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,16384,0.04349866509437561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,51200,0.10759377479553223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,128,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,12288,0.03571910990609063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,10240,0.029875556627909344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,12288,0.01866755551762051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,12288,0.02196888956758711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,10240,0.01923377811908722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,51200,0.13781866762373182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,8192,0.024689777029885188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,10240,0.01701244380739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,10240,0.03164711263444688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,7168,0.021936888496081035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,8192,0.015160888433456421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,8192,0.016362667083740234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,7168,0.013423110875818463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,7168,0.014780445231331719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,6144,0.019194665882322524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,16384,0.022894221875402663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,6144,0.01350400017367469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,5120,0.016367110941145156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,6144,0.022991110881169636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,12288,0.03529599971241421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,65536,0.09126933415730794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,4096,0.013923555612564087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,5120,0.01205599970287747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,5120,0.011448889142937131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,5120,0.02086755633354187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,3584,0.012054222325483957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,4096,0.010599999792046018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,8192,0.027408889598316614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,3072,0.010818666881985135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,4096,0.018947554959191214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,3584,0.010384000009960597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,3584,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,2560,0.009320889082219865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,7168,0.025283555189768474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,3072,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,3584,0.01754399968518151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,3072,0.009397333694828881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,2560,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,2560,0.00905688852071762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,6144,0.012422222230169507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,2560,0.015855999456511605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,1536,0.006799110935793982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,2048,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,2048,0.007847111258241866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,1024,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,2048,0.014646222194035849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,1536,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,1536,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,1536,0.012554667062229581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,768,0.0041964443193541635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,1024,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,1024,0.006032889087994893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,512,0.0032106666929192017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,768,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,4096,0.010843555960390302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,768,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,256,0.002642666714058982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,768,0.010406222608354356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,512,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,512,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,128,0.0022151112142536375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,512,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,256,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,2048,0.007864000068770515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,64,0.002191999927163124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,256,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,3072,0.016784888174798753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,8192,32,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,256,0.009659555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,128,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,65536,0.15586311287350124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,128,0.0074826669361856245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,64,0.005411555369695027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,8192,32,0.00566133318675889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,51200,0.12236977948082818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,51200,0.06551822026570638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,16384,0.0451493329472012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,51200,0.054529776175816856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,16384,0.02604266670015123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,51200,0.1073404418097602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,16384,0.04273600048489041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,12288,0.035750223530663386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,16384,0.021944888763957556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,128,0.0058782220714622075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,10240,0.027776888675159876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,12288,0.020259555843141343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,12288,0.03527911172972785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,65536,0.08629777696397568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,10240,0.017698667115635343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,8192,0.0236488895283805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,10240,0.015843555331230164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,65536,0.06715733475155301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,7168,0.02055199940999349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,1024,0.010694222317801582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,8192,0.014049778381983439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,8192,0.015549333559142219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,8192,0.027504889501465693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,6144,0.017771555317772757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,7168,0.013079110946920184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,65536,0.13413511382208929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,7168,0.02437066700723436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,5120,0.015125332607163323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,6144,0.012079999678664737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,6144,0.012727110750145383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,6144,0.023079999619060095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,5120,0.011065777805116443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,4096,0.013332444760534497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,12288,0.0177831103404363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,5120,0.011500444677140979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,5120,0.02088711162408193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,3584,0.011391110718250275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,4096,0.009768889182143742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,4096,0.01033155537313885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,3072,0.010319111247857412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,4096,0.018795554836591084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,3584,0.008970666262838576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,10240,0.030894223186704848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,2560,0.008997333546479544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,3584,0.009860444400045607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,3584,0.01735911104414198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,3072,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,2048,0.007761778102980719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,3072,0.009199110998047723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,2560,0.00814133303032981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,2560,0.008481777376598781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,1536,0.006498666687144174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,7168,0.013561777770519257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,2560,0.015765332513385348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,2048,0.007280888656775157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,2048,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,1024,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,1536,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,2048,0.014559111661381192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,1536,0.006364444477690591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,768,0.0038853333228164246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,1536,0.013008000122176277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,1024,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,1024,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,512,0.0035617777870761026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,1024,0.011047110789351992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,768,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,768,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,256,0.002542222125662698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,768,0.010411555568377176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,512,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,128,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,512,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,256,0.0058417779703934984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,256,0.00508266646001074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,64,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,256,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,128,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,7168,32,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,3072,0.01682577696111467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,128,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,65536,0.13316267066531712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,64,0.0053297777970631914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,32,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,51200,0.10594311025407578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,65536,0.07852088742785983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,65536,0.06433688931994967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,16384,0.037448889679378934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,51200,0.05943200323316786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,51200,0.05230844351980421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,16384,0.020794666475719877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,16384,0.024382222029897902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,12288,0.029984000656339858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,512,0.0052879999081293745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,65536,0.13400267230139837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,12288,0.017303999927308824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,12288,0.018437332577175565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,12288,0.03494222296608819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,7168,128,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,10240,0.015448000695970325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,10240,0.030564443932639226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,8192,0.0151591118839052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,8192,0.013748444616794586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,7168,0.018582221534517076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,8192,0.026168000366952684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,7168,0.012678222523795234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,7168,0.013060444758998023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,6144,0.01776622235774994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,7168,0.024432000186708238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,6144,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,6144,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,10240,0.02537422213289473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,5120,0.012755555411179861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,6144,0.02234755622016059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,4096,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,5120,0.010421333213647207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,8192,0.020850666695170935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,5120,0.010957332948843638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,10240,0.016314667132165697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,3584,0.009724444813198512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,5120,0.020234665936893888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,4096,0.009684444301658207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,4096,0.00923466682434082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,3072,0.008832889298597971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,4096,0.01826400061448415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,3584,0.009448889229032729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,3584,0.017180444465743173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,3072,0.008919110728634728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,2560,0.007246221933099959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,3072,0.016520889268981088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,2560,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,2560,0.007711999946170383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,2048,0.006671111202902264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,2560,0.015323556131786771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,16384,0.04247377647293938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,1536,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,2048,0.006996444529957241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,2048,0.006996444529957241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,2048,0.01351111133893331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,1536,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,1024,0.00388355553150177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,1536,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,1536,0.01204711116022534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,768,0.00349866681628757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,1024,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,3584,0.008672888908121321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,1024,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,512,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,1024,0.011081777513027191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,768,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,3072,0.007932444413503012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,768,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,256,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,768,0.010018666585286459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,512,0.005962666538026597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,512,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,128,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,512,0.010064000056849586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,256,0.005647999958859549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,256,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,51200,0.10755644241968791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,256,0.009003555609120263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,32,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,128,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,128,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,64,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,65536,0.11383288436465794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,6144,32,0.005487999982304043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,51200,0.09062221977445815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,65536,0.07021511263317533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,65536,0.06203911039564344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,16384,0.03321155574586656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,51200,0.05369955632421705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,51200,0.0503768887784746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,12288,0.02570755614174737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,16384,0.021904889080259535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,16384,0.020398222737842135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,16384,0.042797333664364286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,51200,0.10719199975331624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,10240,0.02174577779240078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,12288,0.01704266667366028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,6144,64,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,128,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,12288,0.034119112624062434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,10240,0.014876445134480795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,10240,0.015134221977657743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,7168,0.01650311052799225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,10240,0.03016266557905409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,8192,0.013167999684810638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,8192,0.013789333403110504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,6144,0.014875556031862894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,8192,0.026073777013354834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,7168,0.012314666476514606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,7168,0.012265777422322167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,5120,0.013110222087966071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,7168,0.024297777149412367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,6144,0.011395555403497485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,6144,0.01149777736928728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,4096,0.011567999919255575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,5120,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,65536,0.1336888869603475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,5120,0.010344888601038191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,3584,0.010228444304731157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,4096,0.009007111191749573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,4096,0.009504000345865885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,12288,0.016747555798954435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,3072,0.009182222187519073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,4096,0.018199111024538677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,8192,0.01854044364558326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,3584,0.008276444342401292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,3584,0.008960000342792934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,2560,0.00792622235086229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,3584,0.01696266730626424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,3072,0.007679999702506595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,3072,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,2048,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,3072,0.016030222177505493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,2560,0.007173333730962541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,2560,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,1536,0.004734222259786394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,2560,0.014645333091417948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,2048,0.0070364442136552595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,2048,0.0063031112982167145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,1024,0.003568888952334722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,2048,0.013430222041077085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,1536,0.00656444455186526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,6144,0.021870222356584337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,1536,0.01240533341964086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,768,0.003176888864901331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,1024,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,1024,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,512,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,5120,0.020454222957293194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,1024,0.010767999622556897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,768,0.006073777874310811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,768,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,256,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,768,0.0103493332862854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,512,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,128,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,512,0.009678222239017487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,256,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,256,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,64,0.0021262222694026097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,128,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,128,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,5120,32,0.002209777840309673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,128,0.00740533322095871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,65536,0.09049422211117214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,64,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,32,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,5120,1536,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,51200,0.07195111115773518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,65536,0.06463911135991414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,65536,0.05988266732957628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,16384,0.027456000447273254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,51200,0.04966310991181267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,51200,0.04827288786570231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,65536,0.1336844497256809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,512,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,12288,0.021354667014545862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,16384,0.019537778364287484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,16384,0.019647111495335896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,51200,0.10648000240325928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,256,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,10240,0.018479110466109384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,16384,0.04198044538497925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,12288,0.01616533266173469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,8192,0.016143111719025504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,12288,0.03357333276006911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,10240,0.0144533332851198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,10240,0.014189332723617554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,7168,0.014257777068350049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,10240,0.029545777373843726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,8192,0.012806221842765808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,8192,0.012735111018021902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,6144,0.01241422196229299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,8192,0.02602400051222907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,7168,0.011687999798191918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,7168,0.01146311147345437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,5120,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,7168,0.02403822210099962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,6144,0.010675555302037133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,6144,0.010835555692513784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,4096,0.009282666775915358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,6144,0.022123555342356365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,5120,0.009408888717492422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,3584,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,5120,0.009630222287442949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,5120,0.019645333290100098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,4096,0.008375999828179678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,3072,0.0069822221994400024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,4096,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,3584,0.008238222036096785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,3584,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,3072,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,3584,0.016467556357383728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,3072,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,2048,0.005395555661784277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,12288,0.016166221764352586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,3072,0.015496888094478183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,2560,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,2560,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,1536,0.004472889006137848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,2048,0.007087111473083496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,2560,0.014160000615649752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,2048,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,1024,0.0035128887328836652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,2048,0.013099555340078143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,1536,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,1536,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,768,0.003088888815707631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,1536,0.01201244443655014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,1024,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,1024,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,512,0.002846222163902389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,1024,0.011053333679835001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,768,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,256,0.002471999989615546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,768,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,768,0.01016266644001007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,512,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,512,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,128,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,4096,0.018014222383499146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,256,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,2560,0.006510222123728857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,256,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,64,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,256,0.008809777597586313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,128,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,128,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,65536,0.07749955521689521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,128,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,64,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,51200,0.061596446567111544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,65536,0.06153866979810926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,65536,0.05894400013817681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,51200,0.047855999734666615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,16384,0.0233608881632487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,51200,0.0473635560936398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,65536,0.13367999924553767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,51200,0.10664266347885132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,16384,0.01939911146958669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,12288,0.01889511115021176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,512,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,16384,0.018930666976504855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,10240,0.016190222567982145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,16384,0.041434665520985924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,12288,0.01586577792962392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,4096,32,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,12288,0.015169777803950839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,8192,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,12288,0.0331911113527086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,10240,0.013736888766288757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,10240,0.014183999763594734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,7168,0.01275022245115704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,10240,0.029100444581773546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,8192,0.012396444049146442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,4096,32,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,8192,0.012549333274364471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,7168,0.01108444482088089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,8192,0.025715554753939312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,7168,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,5120,0.00996800015370051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,7168,0.024021334118313257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,6144,0.010175111393133799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,6144,0.010613333019945355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,4096,0.008954666554927826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,6144,0.02198311189810435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,5120,0.009416888985368941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,3584,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,4096,0.008303999900817871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,4096,0.008672888908121321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,3072,0.007247111035717859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,4096,0.018029333816634286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,3584,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,3584,0.008016889293988546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,2560,0.006467555546098285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,3584,0.01622577839427524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,3072,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,3072,0.007106666763623555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,2048,0.005121777868933148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,3072,0.014782221780882942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,2560,0.007324444750944774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,2560,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,2560,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,6144,0.011710222396585675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,2048,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,2048,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,1024,0.0032417778339650896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,2048,0.01273155543539259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,1536,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,1536,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,768,0.002983111060327954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,1536,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,1024,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,1024,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,5120,0.009515555368529426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,512,0.002761777697337998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,1024,0.010672888822025724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,5120,0.019545776976479422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,768,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,256,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,768,0.010054222411579555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,512,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,512,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,128,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,512,0.009665778113736046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,256,0.00565422202150027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,256,0.004957333207130432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,64,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,256,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,128,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,128,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,65536,0.0683164464102851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,128,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,1536,0.0042035554846127825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,64,0.00519466648499171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,51200,0.05432800120777554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3584,32,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,65536,0.059439109431372754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,65536,0.13307644261254206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,51200,0.04654133319854736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,16384,0.020115555988417733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,51200,0.10654933585060967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,768,0.005975111077229182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,12288,0.016532444291644625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,16384,0.019133332702848647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,16384,0.040764444404178195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,12288,0.014801777071423002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,10240,0.014129777749379476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,12288,0.015795555379655626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,12288,0.03298577666282654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,10240,0.013199110825856527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,10240,0.013855111267831592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,8192,0.01238933371173011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,65536,0.05795377492904663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,10240,0.029116445117526587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,7168,0.011691555380821228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,8192,0.012049777640236748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,51200,0.04600711001290215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,8192,0.012085333466529846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,8192,0.025620443953408137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,6144,0.010656889114114972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,7168,0.011017777853541903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,7168,0.010663111176755695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,5120,0.009205333060688442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,7168,0.023405333360036213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,16384,0.018171555466122098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,6144,0.010223999619483948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,6144,0.021591110361946955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,5120,0.008967111508051554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3584,32,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,5120,0.0196142229768965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,3584,0.0075786668393347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,4096,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,4096,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,3072,0.00693866651919153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,4096,0.01759466694460975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,3584,0.00795555528667238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,3584,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,2560,0.005962666538026597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,3584,0.015816888875431485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,3072,0.007672000262472365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,3072,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,2048,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,3072,0.01443377799457974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,2560,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,2560,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,1536,0.004243555582231945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,2560,0.013468444347381592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,2048,0.007056000332037608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,2048,0.01273155543539259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,6144,0.009716444545321995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,1536,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,1536,0.005958222266700532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,5120,0.008959111240175035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,1536,0.011713777979214987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,768,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,1024,0.006227555374304454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,1024,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,512,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,1024,0.011009777585665384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,768,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,768,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,256,0.0021671110557185277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,768,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,512,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,512,0.0052800000541739995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,128,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,512,0.009986666341622671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,256,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,64,0.0018044445249769424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,256,0.008963555925422246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,128,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,128,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,2048,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,65536,0.05784622165891859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,128,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,64,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,1024,0.0032035555276605818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,3072,32,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,4096,0.008310221963458592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,65536,0.056828445858425565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,65536,0.056277334690093994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,51200,0.04561511013242933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,65536,0.13294933901892766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,51200,0.10691199700037639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,16384,0.01872711049185859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,12288,0.015428443749745687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,16384,0.04050844576623704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,12288,0.014475555883513557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,12288,0.01519466605451372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,10240,0.013584888643688627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,12288,0.03325422273741828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,3072,32,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,8192,0.01143200033240848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,10240,0.013449778159459433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,10240,0.013009777499569787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,51200,0.0450799994998508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,10240,0.029214223225911457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,7168,0.010433778166770935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,16384,0.018206222189797294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,8192,0.01144088887506061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,8192,0.01111288865407308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,6144,0.010289777484205034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,51200,0.044108443790011935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,7168,0.010716444088353051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,7168,0.010900444454616971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,5120,0.009135111338562435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,7168,0.023755555351575215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,6144,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,6144,0.021691555778185528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,5120,0.008966222405433655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,4096,0.007835555407736037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,5120,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,5120,0.019376888871192932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,3584,0.007191999918884701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,4096,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,4096,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,4096,0.016515556308958266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,3072,0.006501333167155583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,3584,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,3584,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,3584,0.015555555621782938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,2560,0.005666666560702854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,3072,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,3072,0.006897777732875612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,8192,0.025621333056026038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,3072,0.014783110883500842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,2048,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,2560,0.007290666302045186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,2560,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,16384,0.017753778232468497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,1536,0.0041999999019834726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,6144,0.009469333622190688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,1024,0.003206222214632564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,1536,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,2048,0.012746666868527731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,1536,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,768,0.002842666581273079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,1536,0.012073777616024017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,1024,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,512,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,768,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,768,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,512,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,512,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,256,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,512,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,256,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,128,0.0019066666977273093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,256,0.005110222018427319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,256,0.008977777428097194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,2048,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,64,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,2048,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2560,32,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,128,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,128,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,128,0.00775022225247489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,65536,0.04681333237224155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,1024,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,64,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,51200,0.03732088870472378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2560,32,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,65536,0.054416000843048096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,768,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,16384,0.015798222687509324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,51200,0.04490400022930569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,51200,0.04251022140185038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,12288,0.013766222529941134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,16384,0.01720711092154185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,51200,0.10613422261344062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,16384,0.018593778212865193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,10240,0.0123831108212471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,16384,0.040488001373079084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,12288,0.014304889572991265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,12288,0.014136888914638095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,8192,0.011384000380833944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,12288,0.03288711110750834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,10240,0.012455110748608908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,10240,0.012730666332774691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,7168,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,10240,0.029109332296583388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,8192,0.01075644459989336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,8192,0.010899555351999072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,6144,0.00963466697269016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,8192,0.02541333271397485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,7168,0.023018666439586218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,65536,0.055757330523596875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,5120,0.008692444198661381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,6144,0.008952889177534316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,6144,0.009653333160612319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,65536,0.13309866852230495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,4096,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,5120,0.008973333570692275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,5120,0.008292444050312042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,5120,0.01824355622132619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,4096,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,3584,0.006677333265542984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,2560,0.013796444568369122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,4096,0.016698666744761996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,3584,0.00777422222826216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,3072,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,3584,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,3584,0.01601777805222405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,7168,0.010119111173682744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,2560,0.005066666752099991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,3072,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,3072,0.014486221803559197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,2560,0.006437333507670297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,2048,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,2560,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,2560,0.01408088869518704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,2048,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,2048,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,1536,0.00388711111413108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,4096,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,2048,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,1024,0.0031635556370019913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,1536,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,1536,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,1536,0.011408888631396823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,768,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,1024,0.00554044461912579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,1024,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,7168,0.009898666706350114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,512,0.002529777793420686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,1024,0.010392000277837118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,768,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,3072,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,768,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,256,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,768,0.009993777506881291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,512,0.0058764442801475525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,6144,0.020608888732062448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,512,0.005283555636803309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,512,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,256,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,256,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,64,0.0019662222928471034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,32,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,256,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,128,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,128,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,65536,0.0355004436439938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,128,0.007621333003044128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,51200,0.029104888439178467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,64,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,2048,32,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,65536,0.054766221178902515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,65536,0.05214133196406894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,16384,0.013817778064144982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,51200,0.04442666636572944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,51200,0.04084622197681003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,65536,0.13275199466281468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,12288,0.011964444484975604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,51200,0.10579288668102688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,16384,0.017495110630989075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,16384,0.016827555166350473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,10240,0.011787555283970304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,12288,0.01347555551264021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,16384,0.04059911105367873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,8192,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,12288,0.013759111364682516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,10240,0.012077333198653327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,12288,0.03288888931274414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,7168,0.00983199973901113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,2048,128,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,8192,0.010757333702511258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,10240,0.02883466747072008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,8192,0.010661332971519895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,6144,0.00923822240697013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,7168,0.00999466660949919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,8192,0.025402666793929204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,7168,0.009682666924264696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,6144,0.009666666388511658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,7168,0.022651554809676275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,6144,0.008923555413881937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,6144,0.020580444071027968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,5120,0.008972444468074376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,5120,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,3584,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,5120,0.019163555569118924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,4096,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,4096,0.0073831114504072405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,3072,0.005671111245950063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,4096,0.016470222009552848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,3584,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,3584,0.015491555134455362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,2560,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,3072,0.006707555717892117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,3072,0.014479999740918478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,10240,0.012055111428101858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,2048,0.004568888909286923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,2560,0.0063475556671619415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,2560,0.007247111035717859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,2560,0.013518222504191928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,5120,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,1536,0.0038417776425679526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,2048,0.006801777829726537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,4096,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,2048,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,1024,0.003157333367400699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,1536,0.00666844430896971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,1536,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,1536,0.011415999796655444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,768,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,1024,0.00620888877246115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,3584,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,1024,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,512,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,1024,0.010480889015727572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,768,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,3072,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,768,0.005286222116814719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,256,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,768,0.010316444767846001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,512,0.005793777604897817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,128,0.0018622221218215094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,512,0.009394666386975182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,256,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,256,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,64,0.001790222194459703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,256,0.009677333136399588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1536,32,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,128,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,2048,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,128,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,64,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,32,0.005085333353943295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,51200,0.023846222294701472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,65536,0.05436088972621494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,65536,0.049455109569761485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,65536,0.13193421893649632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,16384,0.011643555429246692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,51200,0.039552887280782066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,51200,0.10568444596396552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1536,512,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,16384,0.016012445092201233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,12288,0.010855110983053843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,16384,0.01615111033121745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,16384,0.04097155398792691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,10240,0.009696000152164036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,12288,0.013355555633703867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,12288,0.012741333080662621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,65536,0.025999110605981614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,12288,0.03221244282192654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,8192,0.00867911097076204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,10240,0.011417778001891242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,10240,0.012078222301271228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,7168,0.008319999608728621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,10240,0.027857777145173814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,8192,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,8192,0.010045333041085137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,6144,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,8192,0.02444800072246128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,7168,0.009992889232105678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,7168,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,51200,0.04383733206325107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,7168,0.022312889496485393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,6144,0.009668444593747457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,6144,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,4096,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,6144,0.020336000455750358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,5120,0.008637333081828224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,5120,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,4096,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,5120,0.018057778477668762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,4096,0.007268444531493717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,3072,0.005375111268626318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,4096,0.016823111308945548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,3584,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,128,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,2560,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,3584,0.01573422220018175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,3072,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,3072,0.014583110809326172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,2048,0.004316444612211651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,2560,0.006565333240562015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,2560,0.006994666324721442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,2560,0.01384622189733717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,2048,0.00665155549844106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,1536,0.0038773334688610504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,5120,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,2048,0.006246222390068903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,1024,0.0031591111587153543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,2048,0.01276177747382058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,1536,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,1536,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,768,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,3584,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,1024,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,1024,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,512,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,1024,0.010951110886202918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,768,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,3584,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,768,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,256,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,3072,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,512,0.009322666459613377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,256,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,256,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,128,0.0021119999388853707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,256,0.009030222064918941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,128,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,64,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,128,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,1024,32,0.002151111140847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,1536,0.01180355581972334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,65536,0.022966222630606756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,64,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,32,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,768,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,51200,0.018924444913864136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,65536,0.048602667119767934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,65536,0.054210666153166026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,512,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,16384,0.010679999987284342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,65536,0.13217688931359187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,51200,0.04344888859324985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,51200,0.0389013323518965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,12288,0.008967999782827165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,16384,0.015488889482286243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,16384,0.015828443898095023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,16384,0.03973155551486545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,10240,0.008329778081840938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,12288,0.01311822235584259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,128,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,12288,0.012439111040698158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,8192,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,12288,0.03186488813824124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,10240,0.01199288914601008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,10240,0.011425777441925473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,7168,0.007326222128338284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,10240,0.02819733487235175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,8192,0.02403822210099962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,7168,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,1024,512,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,6144,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,7168,0.021980444590250652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,7168,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,5120,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,6144,0.009374221993817223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,6144,0.008967999782827165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,51200,0.1056177748574151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,5120,0.00868711123863856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,6144,0.020375111036830477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,5120,0.0181697772608863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,3584,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,4096,0.008073777788215214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,4096,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,4096,0.016508445143699646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,3072,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,3584,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,3584,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,8192,0.010676444404655032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,2560,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,3072,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,3072,0.006679111056857639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,3072,0.014479110638300577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,2048,0.004247111164861255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,2560,0.007017778025733099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,2560,0.013776889277829064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,2048,0.00610133343272739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,1536,0.003719999972316954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,4096,0.006246222390068903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,5120,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,2048,0.01241422196229299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,1024,0.0032239999208185407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,1536,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,1536,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,1536,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,1024,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,768,0.002801777794957161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,1024,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,8192,0.009948444035318162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,768,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,768,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,768,0.010325333310498131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,2560,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,256,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,512,0.009392889009581672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,2048,0.006824000014199152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,256,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,128,0.001959111127588484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,256,0.005687110953860813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,256,0.008947555389669206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,128,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,128,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,64,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,128,0.007690666450394525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,3584,0.015615110596021017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,65536,0.02015199926164415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,768,512,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,64,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,32,0.004644444419278039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,51200,0.01681066718366411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,65536,0.053391999668545194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,768,512,0.005191111316283544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,512,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,16384,0.009328888522254096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,65536,0.13156178262498644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,51200,0.03859466645452712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,51200,0.03872977693875631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,12288,0.007584888901975419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,16384,0.015855999456511605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,16384,0.015159999330838522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,51200,0.10502844386630589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,10240,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,16384,0.03935822182231479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,12288,0.012344888514942594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,12288,0.013425778183672162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,8192,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,10240,0.011941333611806234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,10240,0.011115555961926779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,1024,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,7168,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,8192,0.01074222226937612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,8192,0.010021333065297868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,10240,0.027465777264700994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,6144,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,8192,0.023800000548362732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,7168,0.01016622202263938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,65536,0.048244443204667836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,7168,0.009341333475377824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,5120,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,6144,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,6144,0.020057777563730877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,5120,0.008965333302815756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,4096,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,5120,0.018231110440360177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,3584,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,4096,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,4096,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,4096,0.016559110747443307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,3072,0.005439110928111606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,12288,0.031215998861524794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,3584,0.0079297779334916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,3584,0.015138667490747241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,2560,0.004859555512666702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,3072,0.00758755538198683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,3072,0.006893333461549547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,3072,0.01441511180665758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,2048,0.0042364444169733255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,2560,0.006439111298984951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,2560,0.0070426662762959795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,6144,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,1536,0.003540444291300244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,2048,0.006948444578382704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,2048,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,5120,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,2048,0.01221244451072481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,1024,0.002846222163902389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,1536,0.006325333482689328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,1536,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,768,0.002834666727317704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,1536,0.011815999945004782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,1024,0.006057777752478917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,1024,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,512,0.00254488881263468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,1024,0.010363555616802638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,3584,0.007187555233637492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,768,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,256,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,768,0.010412444670995077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,512,0.0057857777509424425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,512,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,128,0.0017911110901170308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,512,0.009364444348547194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,256,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,256,0.009368889033794403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,128,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,2560,0.013783111340469785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,128,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,32,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,65536,0.012824888858530255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,128,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,64,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,512,32,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,51200,0.011379555695586734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,65536,0.045528001255459256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,65536,0.046579556332694165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,65536,0.131221334139506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,16384,0.0063884444534778595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,768,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,51200,0.03748177819781833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,51200,0.10502311256196763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,16384,0.015777778294351365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,12288,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,16384,0.014827556080288358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,256,0.00564533347884814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,16384,0.03903288973702325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,10240,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,12288,0.01310400002532535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,12288,0.01238933371173011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,512,64,0.001814222170246972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,7168,0.021798221601380244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,8192,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,10240,0.011758222348160215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,10240,0.027506666051016912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,8192,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,7168,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,8192,0.010552888943089379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,8192,0.02437688906987508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,6144,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,7168,0.009768000079525841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,7168,0.00923911150958803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,7168,0.021681777305073206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,51200,0.03616088959905837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,5120,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,6144,0.00868711123863856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,6144,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,4096,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,6144,0.020240889655219186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,5120,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,5120,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,3584,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,5120,0.017542221479945712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,4096,0.008117333054542542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,4096,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,3072,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,4096,0.016201777590645682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,12288,0.03165600034925673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,3584,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,3584,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,10240,0.011403555671374003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,3072,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,3584,0.01514844430817498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,3072,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,2048,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,3072,0.014427555931939019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,2560,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,2560,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,1536,0.003560888684458203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,2560,0.013501333693663279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,2048,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,1024,0.002967111145456632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,1536,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,2048,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,768,0.0028106667515304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,1536,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,1024,0.006226666685607698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,1024,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,512,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,1024,0.011033777561452655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,768,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,768,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,256,0.0021520000365045336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,768,0.009957333405812582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,512,0.0058782220714622075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,512,0.005119110975000593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,128,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,512,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,256,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,256,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,2560,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,64,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,256,0.008632000121805403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,128,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,128,0.004855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,128,0.007677333222495184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,64,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,32,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,51200,0.01033333357837465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,256,2048,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,65536,0.0461368891927931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,1536,0.011289777855078379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,51200,0.03737066520584954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,65536,0.13093333774142796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,51200,0.03539289037386576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,51200,0.10388800170686509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,12288,0.005295111073387994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,16384,0.015819556183285184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,16384,0.014881778094503613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,16384,0.03967644439803229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,10240,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,12288,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,12288,0.012156444291273752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,256,32,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,12288,0.031207111146714952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,8192,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,65536,0.011731555892361535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,10240,0.011398221883508893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,7168,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,10240,0.028149333265092637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,8192,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,8192,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,6144,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,65536,0.044172445933024086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,7168,0.010047111246320937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,7168,0.009353777600659264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,5120,0.004944889081848992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,16384,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,6144,0.009330666727489894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,6144,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,6144,0.01995822290579478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,4096,0.004961777892377642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,5120,0.008661333057615492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,5120,0.007786666353543599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,5120,0.018529777725537617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,3584,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,4096,0.007954667011896769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,4096,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,4096,0.015951999359660678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,3072,0.004516444272465176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,10240,0.011759111450778114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,3584,0.007251555720965068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,3584,0.015166223049163818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,2560,0.004574222283230888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,3072,0.00740622232357661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,3072,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,8192,0.02380888826317257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,3072,0.014190221826235453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,2048,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,2560,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,7168,0.022293332550260756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,2560,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,1536,0.003522666792074839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,2048,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,2048,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,1024,0.002838222309947014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,1536,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,2048,0.012079999678664737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,1536,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,768,0.0025395556456512874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,1536,0.012159999873903064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,1024,0.005984000033802456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,1024,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,512,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,1024,0.010332444475756751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,3584,0.0076248885856734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,768,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,256,0.0021262222694026097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,768,0.010703110860453712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,512,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,512,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,128,0.0019991110182470745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,512,0.009291555318567488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,256,0.0053084443012873335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,256,0.004858666823969947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,64,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,256,0.00965244405799442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,128,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,128,32,0.001812444378932317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,128,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,65536,0.009319110876984065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,2560,0.014025777578353882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,51200,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,128,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,12288,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,32,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,10240,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,65536,0.04424622323777941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,51200,0.0357582237985399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,16384,0.01499466598033905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,12288,0.012111999922328524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,7168,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,6144,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,10240,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,768,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,5120,0.004550222307443619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,8192,0.00998755544424057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,4096,0.004295999805132548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,7168,0.008967111508051554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,3584,0.004856889032655292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,3072,0.004559111264016894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,6144,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,2560,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,5120,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,2048,0.00424888895617591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,4096,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,3584,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,1536,0.0035173334181308746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,1024,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,2560,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,768,0.002842666581273079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,2048,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,16384,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,1536,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,128,64,0.004535111288229625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,256,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,1024,0.005325333525737126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,128,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,64,0.0017431111385424931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,768,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,32,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,8192,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,512,0.005040888984998067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,65536,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,256,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,128,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,51200,0.00902844468752543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,16384,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,64,0.004586666822433472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,12288,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,32,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,10240,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,65536,0.04380711250834995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,51200,0.03542933199140761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,8192,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,16384,0.01447111037042406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,12288,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,6144,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,10240,0.011156444748242697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,5120,0.004570666700601578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,8192,0.009659555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,64,3072,0.0069057775868309875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,4096,0.004212444441186057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,7168,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,3584,0.004214222232500712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,3072,0.0041946665280395085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,6144,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,2560,0.004360888981156879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,64,512,0.0022479999396536085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,5120,0.008177777959240807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,2048,0.004228444563017951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,4096,0.007584888901975419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,1536,0.0032479998966058097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,1024,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,3072,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,768,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,2560,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,512,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,2048,0.006333333336644703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,1536,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,256,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,128,0.0018719999740521114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,1024,0.005314666777849197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,64,0.0014888888431919946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,768,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,32,0.0017928888814316855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,512,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,256,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,48,32,7168,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,64,0.004591111093759537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,32,0.00452533322903845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,16384,0.368392891354031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,16384,0.16523999638027614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,3584,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,51200,0.4926026662190755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,12288,0.2724711100260417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,16384,0.13239110840691462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,10240,0.22716977861192492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,12288,0.12325955761803521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,51200,1.1307493845621746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,12288,0.09938400321536595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,8192,0.18557600180308023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,48,32,128,0.004982222285535601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,10240,0.08443999952740139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,7168,0.16267377800411648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,8192,0.08870489067501491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,6144,0.13746221860249838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,7168,0.06113066938188341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,8192,0.07592266798019409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,7168,0.0746613343556722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,6144,0.06568444437450834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,6144,0.053912887970606484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,4096,0.09337333175871108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,5120,0.05564888980653551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,5120,0.05997244517008463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,3584,0.08358222246170044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,4096,0.03892177674505446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,4096,0.04678488771120707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,3072,0.0715235537952847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,3584,0.034832000732421875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,2560,0.05976800123850504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,3584,0.04898044466972351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,2048,0.048103110657797925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,3072,0.030164443784289893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,3072,0.045891556474897593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,1536,0.03948533203866746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,2560,0.03273511264059279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,2560,0.03684355484114753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,10240,0.10468977689743042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,2048,0.022553778356975977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,1024,0.0277502222193612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,2048,0.03122755553987291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,768,0.020255110330051847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,1536,0.018519111805491976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,51200,0.4271093209584554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,1024,0.014675555957688225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,1536,0.02529955572552151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,768,0.013335111240545908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,5120,0.11603821648491754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,768,0.014577777849303352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,512,0.011964444484975604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,128,0.005405333307054307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,256,0.009104889300134447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,512,0.011798222031858234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,64,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,256,0.010509333676762052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,128,0.008263111114501953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,64,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,32,0.012730666332774691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,512,0.013559111290507846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,65536,1024,0.017987555927700467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,256,0.008356444537639618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,65536,0.5214444266425239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,65536,128,0.010655110908879174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,65536,32,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,65536,1.0760364532470703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,51200,0.3947430981530084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,16384,0.27235910627577037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,16384,0.13257243898179796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,51200,0.842999087439643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,12288,0.2050293286641439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,12288,0.09951288832558526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,16384,0.10691022210650974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,12288,0.08135466443167792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,10240,0.08441599872377183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,10240,0.17179021570417616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,51200,0.33393510182698566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,8192,0.13951555887858072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,65536,0.4110551145341661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,7168,0.12270577748616536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,10240,0.06825333171420626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,8192,0.07176444265577528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,6144,0.10442133082283868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,6144,0.053364445765813194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,7168,0.060885330041249595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,7168,0.050844444168938525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,4096,0.07150044706132677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,3584,0.06329688760969374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,5120,0.04549244377348158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,5120,0.04846133457289802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,4096,0.032680001523759626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,4096,0.03751555416319106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,3072,0.055580443806118436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,2560,0.046104888121287026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,3584,0.040004445446862116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,3072,0.03573066658443875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,2048,0.0372631119357215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,2560,0.022825777530670166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,3072,0.02575911084810893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,2560,0.03090311090151469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,1536,0.029137776957617864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,2048,0.019301333361201815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,1024,0.021020443903075323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,1536,0.01614755557643043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,2048,0.025781333446502686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,768,0.015390222271283468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,1536,0.01958311100800832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,5120,0.088664001888699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,6144,0.04381600022315979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,1024,0.013439111411571503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,512,0.010665777656767102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,256,0.006810666786299811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,1024,0.014835556348164877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,768,0.011910222470760345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,128,0.0038702223036024305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,512,0.010469333165221743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,768,0.012449777788586087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,64,0.0032213332338465583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,512,0.010693333215183683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,51200,32,0.003536888708670934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,256,0.009356444080670675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,3584,0.028934223784340754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,128,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,256,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,64,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,51200,32,0.010783111055692038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,65536,0.3617662323845758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,51200,0.28369421429104275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,65536,0.1791048844655355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,51200,0.13309689362843832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,16384,0.07997600237528484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,65536,0.14314844873216417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,12288,0.061724444230397545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,16384,0.044622222582499184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,51200,0.1144791046778361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,16384,0.04797244403097364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,12288,0.03251288996802436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,8192,0.04311555624008179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,12288,0.03708622190687392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,10240,0.02847911251915826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,7168,0.03659288750754462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,10240,0.03199555476506551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,8192,0.025616889198621113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,128,0.009650666680600908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,6144,0.03180177675353156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,7168,0.02167911165290409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,5120,0.027098667290475633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,7168,0.024143111374643113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,6144,0.019545776976479422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,6144,0.021691555778185528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,5120,0.021070222059885662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,3584,0.019799111617936026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,5120,0.018917333748605516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,4096,0.015213333898120456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,4096,0.015803555647532146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,51200,8192,0.05681066380606758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,3072,0.017630222770902846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,3584,0.017607111069891188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,2560,0.014793778459231058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,3072,0.013233777549531726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,10240,0.05067200130886502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,2048,0.012359999948077731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,2560,0.01204711116022534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,2560,0.013882666826248169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,1536,0.009906666974226633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,2048,0.010785777535703448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,8192,0.027710222535663184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,2048,0.012270222107569376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,1024,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,1536,0.010652444428867765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,4096,0.02272266646226247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,768,0.006127111199829314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,1536,0.009311999711725447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,1024,0.008486222061845992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,512,0.004369777937730153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,768,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,256,0.002814222127199173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,512,0.006754666566848755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,3584,0.01696622206105126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,512,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,128,0.0025191110455327565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,256,0.006711999989218182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,64,0.0022008888837363985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,256,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,3072,0.015786666009161208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,16384,32,0.0021911110315057966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,128,0.005391110976537068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,64,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,32,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,65536,0.26655999819437665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,65536,0.14938044548034668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,51200,0.20862132973141143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,1024,0.007690666450394525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,65536,0.1328764491611057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,16384,0.07335733042822944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,51200,0.11446666717529297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,12288,0.05673333009084066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,16384,768,0.007322666545708974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,16384,0.041617777612474226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,10240,0.04597244328922696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,16384,0.04192799992031521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,8192,0.037661333878835045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,12288,0.02968977888425191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,16384,128,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,7168,0.03199999862247043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,10240,0.025648888614442613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,10240,0.028082665469911363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,6144,0.02788355615403917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,8192,0.02169955604606205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,8192,0.024489776955710515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,7168,0.019706666469573975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,6144,0.017959111266665988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,4096,0.020227554771635268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,6144,0.019168888529141743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,5120,0.01887022289964888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,5120,0.016511999898486666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,3584,0.017483555608325534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,4096,0.014239999983045789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,3072,0.015449777245521545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,3584,0.01480977733929952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,3584,0.015541333291265698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,2560,0.012991111311647626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,3072,0.012438221938080259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,51200,0.10823822021484375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,3072,0.014296889305114746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,2048,0.011017777853541903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,2560,0.011305777562989129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,12288,0.03252089023590088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,2560,0.01275288893116845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,2048,0.011249778171380361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,1024,0.007376000285148621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,5120,0.0239991115199195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,1536,0.009637333452701569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,768,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,7168,0.021408889028761122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,1024,0.007357333269384172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,1024,0.0074471111098925276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,512,0.004212444441186057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,768,0.0074337778819931885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,256,0.003198222153716617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,768,0.006817777951558431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,512,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,4096,0.0144195556640625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,128,0.0028106667515304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,512,0.006136888845099344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,256,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,64,0.002471999989615546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,32,0.002549333290921317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,128,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,128,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,64,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,32,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,12288,1536,0.00905955582857132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,65536,0.23084266980489096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,1536,0.00832622249921163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,51200,0.18034666114383271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,65536,0.14167111449771458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,16384,0.0591813325881958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,65536,0.12383555041419135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,51200,0.1091466678513421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,12288,0.04661155409283108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,51200,0.10109866989983453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,10240,0.039502223332722984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,16384,0.04003377755482992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,12288,0.027699554959932964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,8192,0.03402400016784668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,12288,256,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,10240,0.026817777090602454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,7168,0.02979822291268243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,8192,0.02072355482313368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,8192,0.023005333211686876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,6144,0.025633777181307476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,12288,2048,0.00983911090426975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,7168,0.018965333700180054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,5120,0.021925333473417494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,6144,0.01722044414944119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,6144,0.017874666386180453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,4096,0.018917333748605516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,5120,0.017830222845077515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,16384,0.03784888982772827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,4096,0.01408266690042284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,3584,0.016248888439602323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,12288,0.03089244498146905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,3072,0.014256000518798828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,10240,0.025106666816605464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,3584,0.012667555775907306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,2560,0.012196444802814059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,3584,0.013812444276279874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,3072,0.011875555747085147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,2048,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,3072,0.012625777886973487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,2560,0.011046222514576383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,1536,0.008660444782839881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,2560,0.011361777782440186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,7168,0.020218667056825426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,1024,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,2048,0.010218666659461128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,1536,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,1536,0.008636444807052612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,768,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,1024,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,512,0.00423200014564726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,1024,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,5120,0.015831111205948722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,768,0.0069439998931354946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,256,0.0032400000426504347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,128,0.0028071111689011254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,512,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,768,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,64,0.002505777817633417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,256,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,512,0.005960000058015187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,10240,32,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,128,0.0063182223174307085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,256,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,128,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,64,0.005777777896987067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,65536,0.1767617728975084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,10240,32,0.006311111152172089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,51200,0.13874755965338811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,16384,0.04641866683959961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,65536,0.09150222274992202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,12288,0.0358915560775333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,2048,0.009150221943855286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,51200,0.07072177860471937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,51200,0.05589866638183594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,16384,0.028155555327733357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,10240,0.029502222935358684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,8192,0.025578666064474318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,12288,0.018403554956118267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,12288,0.021937777598698933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,10240,0.016182222300105624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,7168,0.02160800000031789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,10240,0.01924088928434584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,8192,0.014488889111412896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,6144,0.018927110566033255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,8192,0.017217778497272067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,7168,0.013330666555298699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,5120,0.016146666473812528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,10240,4096,0.013815999858909182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,7168,0.01480977733929952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,6144,0.012429333395428128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,4096,0.013960889644092984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,3584,0.012076444096035428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,5120,0.013077333569526672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,6144,0.013416888813177744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,3072,0.010664888554149203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,4096,0.010445333189434476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,4096,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,2560,0.009402666654851701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,3072,0.008895999855465358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,3584,0.011015999648306104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,65536,0.07169689072502984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,2048,0.0079644446571668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,3072,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,1536,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,2048,0.007343110938866933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,2560,0.009475555684831407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,2048,0.008513777620262569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,1024,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,1536,0.006973333656787872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,1536,0.006710222197903528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,768,0.00388711111413108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,1024,0.006315555423498154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,1024,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,512,0.002838222309947014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,768,0.005982222242487802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,768,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,256,0.002492444382773505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,512,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,5120,0.012034667034943899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,512,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,128,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,3584,0.009333333207501305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,64,0.002128888956374592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,256,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,128,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,8192,32,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,128,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,2560,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,8192,16384,0.022671999202834234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,65536,0.15732799635993108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,64,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,8192,32,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,16384,0.043212443590164185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,51200,0.12422222561306423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,65536,0.08380977974997626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,12288,0.033484445677863225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,65536,0.06637777884801228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,10240,0.028492444091373022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,51200,0.06455466482374403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,51200,0.05418755610783895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,16384,0.02576711111598545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,8192,0.024079110887315538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,12288,0.017841777867741056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,10240,0.015823110938072205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,10240,0.017525333497259352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,8192,0.013947554760509066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,6144,0.017870222528775532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,8192,0.015784000356992085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,7168,0.012977778083748288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,7168,0.013778666655222574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,5120,0.015251555376582675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,6144,0.012116444607575735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,4096,0.0129457778400845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,6144,0.012773333324326409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,5120,0.011029333704047732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,3584,0.011397333608733283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,5120,0.011391110718250275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,4096,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,3072,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,4096,0.010280888941552904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,3584,0.0086986662613021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,2560,0.008943110704421997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,3072,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,3072,0.010419555836253695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,16384,0.021887999441888597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,2048,0.00775822252035141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,1536,0.006330666856633292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,2560,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,12288,0.020253333780500624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,2560,0.009304000271691216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,1024,0.004604444321658877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,7168,0.02049688829316033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,2048,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,768,0.003950222084919612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,1536,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,2048,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,512,0.003160888950030009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,1024,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,1536,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,256,0.002505777817633417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,1024,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,128,0.0022124445272816554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,768,0.00562755556570159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,512,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,64,0.0021564445147911706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,256,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,512,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,128,0.00563822231358952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,128,0.005155555489990446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,64,0.0052844443255000645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,65536,0.13685688707563612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,32,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,3584,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,65536,0.07617421944936116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,65536,0.06416000260247125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,16384,0.0356026656097836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,51200,0.05825510952207777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,51200,0.05201955636342367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,16384,0.021010667085647583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,16384,0.024104000793562994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,10240,0.02498577700720893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,7168,768,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,12288,0.01719999975628323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,8192,0.02071377800570594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,10240,0.015199111567603217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,7168,32,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,7168,256,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,10240,0.016181333197487723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,7168,0.018647111124462552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,8192,0.013660444153679742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,6144,0.01683911184469859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,8192,0.014648889501889547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,7168,0.012718222207493253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,5120,0.014602666099866232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,7168,0.013038222160604266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,51200,0.09640622138977051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,6144,0.011732444167137146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,4096,0.011686222420798408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,5120,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,5120,0.010791111323568555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,12288,0.028869334194395278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,4096,0.0098213329911232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,4096,0.008960888617568547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,3072,0.009642666412724389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,3584,0.008562666674455007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,2560,0.008272888759771982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,3072,0.007954667011896769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,12288,0.018428444862365723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,3072,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,2048,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,2560,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,1536,0.005816000203291575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,2048,0.00703022215101454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,2560,0.008263111114501953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,1024,0.0037866665257347953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,2048,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,1536,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,768,0.003474666840500302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,1536,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,1024,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,512,0.0028106667515304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,1024,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,768,0.006053333481152852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,256,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,768,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,512,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,6144,0.012063110868136087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,128,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,3584,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,64,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,256,0.005646222167544895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,6144,32,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,6144,128,0.0058702222175068325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,128,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,3584,0.01054844425784217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,64,0.0052800000541739995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,32,0.005445333404673471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,51200,0.08637422323226929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,16384,0.032030221488740705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,65536,0.0617484450340271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,65536,0.0698631140920851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,12288,0.02568799919552273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,51200,0.050105777051713735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,16384,0.02017155620786879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,10240,0.020953777763578627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,16384,0.021765333082940843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,12288,0.016709332664807636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,12288,0.01683911184469859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,8192,0.017625777257813346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,10240,0.01481244464715322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,7168,0.016116445263226826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,10240,0.015030221806632148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,512,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,8192,0.013108444710572561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,6144,0.014104889498816596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,7168,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,6144,256,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,5120,0.013968888256284924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,7168,0.012123555772834353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,6144,0.011180444724029966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,4096,0.011422221859296164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,6144,0.011427555647161273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,5120,0.010001777774757808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,3584,0.01000533335738712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,5120,0.010332444475756751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,4096,0.008660444782839881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,3072,0.008985777695973715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,3584,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,4096,0.009314667019579146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,2560,0.007679999702506595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,3584,0.00997866690158844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,2048,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,51200,0.053004443645477295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,3072,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,1536,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,2560,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,2048,0.007070221834712558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,1024,0.0035457776652442086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,2048,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,1536,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,768,0.0031662223239739737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,1536,0.006339555399285422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,1024,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,512,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,1024,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,768,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,768,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,256,0.002491555487116178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,512,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,65536,0.10787733395894368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,128,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,512,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,256,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,64,0.0021679999513758556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,256,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,128,0.0057297779454125305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,5120,32,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,3072,0.007927999728255803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,65536,0.08624888790978326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,5120,2560,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,64,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,32,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,51200,0.06952977842754789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,16384,0.025967111190160114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,65536,0.05957511398527357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,51200,0.048073778549830117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,51200,0.048672888014051646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,12288,0.021288888321982488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,16384,0.019552888141738046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,8192,0.013879111243618859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,10240,0.018575110369258456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,16384,0.01935822268327077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,12288,0.016136000553766888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,8192,0.01663466625743442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,10240,0.01442577772670322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,7168,0.01443911095460256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,8192,0.012755555411179861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,8192,0.01275022245115704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,6144,0.011642666326628791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,7168,0.011521778172916837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,7168,0.011426666544543372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,5120,128,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,5120,0.010023111270533668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,6144,0.010495111346244812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,4096,0.00945688866906696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,5120,0.008850666383902231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,5120,0.009864889085292816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,65536,0.0635697775416904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,4096,0.008278222547637092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,4096,0.008652444514963362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,3072,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,3584,0.007853333320882585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,3584,0.008071111308203803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,2560,0.006395555618736479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,3072,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,12288,0.015497777197096082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,3072,0.007819555699825287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,2048,0.0053031109273433685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,1536,0.0041964443193541635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,2560,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,10240,0.014030221435758801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,1024,0.003272888975010978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,2560,0.007127111156781514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,2048,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,768,0.0031893334041039148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,2048,0.006657777975002925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,1536,0.006310222049554189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,512,0.0028560000161329904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,1024,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,1536,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,256,0.0024577778660588795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,768,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,1024,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,512,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,768,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,6144,0.010729778144094678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,128,0.0022168887986077205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,64,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,512,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,256,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,32,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,65536,0.07350844144821167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,4096,3584,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,128,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,64,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,51200,0.05858666367001004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,32,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,16384,0.024248000648286607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,65536,0.06048266755210029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,65536,0.058740443653530545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,12288,0.019351111518012155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,51200,0.047114666965272695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,16384,0.01925155520439148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,51200,0.046795556942621865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,16384,0.018636445204416912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,12288,0.015840889679061044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,8192,0.01460533340771993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,12288,0.01497688889503479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,10240,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,7168,0.013431111143694984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,8192,0.012233778006500669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,10240,0.013495999905798169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,6144,0.01262399968173769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,8192,0.01238222254647149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,7168,0.01127644462717904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,5120,0.01126577787929111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,7168,0.011014222270912595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,6144,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,6144,0.010608889162540436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,5120,0.009015999734401703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,4096,128,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,3584,0.007677333222495184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,4096,0.008492444124486711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,4096,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,3072,0.0069013333155049225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,3584,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,3584,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,2560,0.006216889040337668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,10240,0.01648533344268799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,3072,0.007711999946170383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,3072,0.007238222493065729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,2048,0.005122666557629903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,2560,0.0074480002125104266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,2048,0.006986666884687211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,2048,0.006565333240562015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,1536,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,1536,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,4096,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,1024,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,768,0.002874666617976295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,1024,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,768,0.006063111126422882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,512,0.0025519999778932994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,768,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,512,0.00573600000805325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,256,0.002493333278430833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,512,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,2560,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,256,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,128,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,256,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,64,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3584,128,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,32,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,128,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,65536,0.06547200017505221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,5120,0.009383999639087254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,32,0.0052897776994440295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,51200,0.05229599939452278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,65536,0.05760711431503296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,65536,0.05707910988065931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,16384,0.020388444264729817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,51200,0.04620711008707682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,51200,0.04516355527771843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,12288,0.01687111126052009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,16384,0.018952889574898612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,10240,0.014536889062987434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,12288,0.015825778245925903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,1536,0.004122666600677702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,12288,0.014705777168273926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,8192,0.012616888516479068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,10240,0.013827555709415011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,7168,0.011759111450778114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,10240,0.013263999587959714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,8192,0.011762667033407422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3584,1024,0.003228444399105178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,6144,0.010619555910428366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,8192,0.011944889194435544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,7168,0.010601777997281818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,5120,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,7168,0.01092177795039283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3584,64,0.004990222139490975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,6144,0.009650666680600908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,4096,256,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,3584,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,5120,0.00905955582857132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,6144,0.010155555274751451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,3072,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,4096,0.008352888955010308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,4096,0.00755555596616533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,3584,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,2560,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,3072,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,2048,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,3072,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,2560,0.007252444823582967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,1536,0.0038559999730851916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,2560,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,16384,0.017831999394628737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,2048,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,1024,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,2048,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,1536,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,1024,0.006227555374304454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,1024,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,512,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,768,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,768,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,256,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,512,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,4096,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,128,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,512,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,256,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,5120,0.008893333375453949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,256,0.00507733349998792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,64,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,3072,128,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,32,0.0018817778262827133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,128,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,3584,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,64,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,32,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,51200,0.047334220674302846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,65536,0.05643466446134779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,16384,0.01864977843231625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,65536,0.055325332615110606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,3072,768,0.0028764444092909494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,51200,0.04522666666242811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,12288,0.015788444214397006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,16384,0.018802666001849704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,10240,0.013724444641007317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,16384,0.017553778158293832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,12288,0.01553244392077128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,8192,0.01202755586968528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,12288,0.014466666513019137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,10240,0.012863111164834766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,7168,0.011387555963463254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,8192,0.011039999624093374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,8192,0.011600000162919363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,6144,0.010267555713653564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,7168,0.01036977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,65536,0.055452442831463285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,7168,0.01071911139620675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,5120,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,6144,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,4096,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,6144,0.009323555562231276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,5120,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,3584,0.007206222249401941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,5120,0.00850399997499254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,4096,0.008413333031866286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,3072,0.00656088896923595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,3072,1536,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,4096,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,51200,0.04335022303793165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,3584,0.007802666889296637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,2560,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,2048,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,3584,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,3072,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,2560,0.007332444190979004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,1536,0.003843555433882607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,3072,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,2048,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,1024,0.0031840000301599503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,2560,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,2048,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,1536,0.006511111226346757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,1024,0.006305777778228124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,1536,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,512,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,768,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,1024,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,256,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,768,0.005474666754404704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,512,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,128,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,256,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,512,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,64,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,32,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,256,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,128,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,65536,0.04902844296561348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,51200,0.0383200016286638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,32,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,65536,0.05287999908129374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,16384,0.016172443826993305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,51200,0.04481155673662821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,51200,0.04157244496875339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,10240,0.012792000340090858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,12288,0.013567999833159976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,16384,0.01833777791923947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2560,768,0.002797333316670524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,16384,0.016968000266287062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,12288,0.014127999544143677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,8192,0.011326221956147088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,10240,0.012255999777052136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,12288,0.013904000322024027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,7168,0.010196444061067369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,10240,0.012715555727481842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,8192,0.01074399964676963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,6144,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,7168,0.010001777774757808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2560,128,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,7168,0.009759999811649323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,5120,0.008836444053384993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,6144,0.009639999932712978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2560,64,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,4096,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,5120,0.009034666750166152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,3584,0.006897777732875612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,65536,0.05548888776037428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,5120,0.008240888516108194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,4096,0.008080888953473832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,3072,0.005728000154097875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,4096,0.0074275558193524676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,3584,0.007132444116804335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,2560,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,3072,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,10240,0.012157333393891653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,3072,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,2048,0.005537777725193236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,2560,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,1536,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,2560,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,2048,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,1024,0.003167111012670729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,1536,0.006663110935025745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,8192,0.010915555887752108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,768,0.0029013332807355454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,1536,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,1024,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,512,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,1024,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,768,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,6144,0.009157333109113906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,768,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,512,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,128,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,256,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,3584,0.007730666961934831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,256,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,64,0.0018746666610240936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,128,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,32,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,128,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,65536,0.03709422217475043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,64,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,2048,2048,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,51200,0.029461334149042766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,65536,0.05060355530844795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,16384,0.013417777915795645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,51200,0.044334222873051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,51200,0.040038221412234835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,12288,0.012106666962305704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,16384,0.017248888810475666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,16384,0.016586666305859882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,10240,0.010739555789364709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,2048,256,0.0021608887861172357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,12288,0.013528889252079858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,8192,0.009673777553770278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,12288,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,10240,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,512,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,7168,0.009115555220180089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,10240,0.011975111232863532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,6144,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,8192,0.010672888822025724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,7168,0.010031999813185798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,5120,0.008651555412345463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,7168,0.009666666388511658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,6144,0.009322666459613377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,4096,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,6144,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,2048,32,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,65536,0.05475733346409268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,5120,0.008184000021881526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,3072,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,4096,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,3584,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,2560,0.005140444470776452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,3584,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,3072,0.007418666448858049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,2048,0.004210666649871402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,3072,0.006908444480763541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,2560,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,2560,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,2048,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,2048,0.006081777728266186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,1024,0.0031600000543726813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,1536,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,1536,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,768,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,1024,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,8192,0.01032977799574534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,1024,0.005777777896987067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,512,0.0025271111064487034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,768,0.005966222120655908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,768,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,512,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,5120,0.008989333278603023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,3584,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,128,0.0018639999131361644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,512,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,64,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,128,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,256,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,32,0.0018035556293196147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,128,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,65536,0.028351111544503108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,64,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,51200,0.022610666023360357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,1536,0.003834666477309333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1536,32,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,65536,0.05452533231841194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,65536,0.048455112510257296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,51200,0.043819556633631386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,51200,0.038880000511805214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,12288,0.009302222066455418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,16384,0.01611911091539595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,12288,0.013232000172138214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1536,256,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,10240,0.008585777547624376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,12288,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,4096,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1536,256,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,10240,0.01204355557759603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,7168,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,10240,0.011347555451922946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,8192,0.01071111112833023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,8192,0.009938666390048133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,6144,0.006724444528420766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,7168,0.010090666512648264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,5120,0.00830844458606508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,7168,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,6144,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,4096,0.007099555598364935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,16384,0.011401777466138205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,6144,0.008712000317043727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,5120,0.008926221893893348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,3584,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,5120,0.008153777983453538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,4096,0.008006222546100616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,3072,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,3584,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,4096,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,16384,0.016144000821643405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,2560,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,3584,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,2560,0.006973333656787872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,2560,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,1536,0.0038720000949170855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,8192,0.007816889219813878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,2048,0.006767999794748094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,2048,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,1024,0.003175111073586676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,1536,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,768,0.002874666617976295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,1536,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,1024,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,512,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,768,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,768,0.005296000176005893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,256,0.002201777779393726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,512,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,512,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,128,0.001960888918903139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,256,0.005490666462315454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,64,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,256,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,32,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,1024,2048,0.004386666748258802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,128,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,65536,0.02328088879585266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,128,0.004982222285535601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,3072,0.006821333534187741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,51200,0.018984888990720112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,64,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,16384,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,65536,0.05416711171468099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,1024,32,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,65536,0.04747822218471103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,12288,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,51200,0.038559112283918597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,16384,0.015837333268589444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,1024,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,12288,0.013067555924256643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,16384,0.014869333969222175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,8192,0.009647111097971598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,12288,0.01239822225438224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,10240,0.011745777395036487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,7168,0.008641777767075432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,8192,0.010492444038391113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,10240,0.011408000356621213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,6144,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,8192,0.00979377743270662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,7168,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,5120,0.007177777588367462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,6144,0.009361777868535783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,7168,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,4096,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,6144,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,3584,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,5120,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,4096,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,1024,3072,0.007290666302045186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,3072,0.0053084443012873335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,4096,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,2560,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,3072,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,3584,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,2048,0.004283555679851108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,3072,0.006688888702127669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,51200,0.043615112702051796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,2560,0.007075555622577667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,2560,0.00630044440428416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,10240,0.010961777634090848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,2048,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,1024,0.003152888889114062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,2048,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,1536,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,1536,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,768,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,1024,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,512,0.002505777817633417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,1024,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,768,0.005967999911970562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,256,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,768,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,512,0.00573511090543535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,5120,0.008678221868144141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,128,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,512,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,64,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,256,0.005290666802061929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,32,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,128,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,256,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,65536,0.016576889488432143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,768,3584,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,128,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,51200,0.01370133293999566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,64,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,16384,0.007807999849319458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,65536,0.05262400044335259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,51200,0.03873422079616123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,768,1536,0.003666666646798452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,12288,0.009639999932712978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,51200,0.03790488839149475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,16384,0.01588088936275906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,10240,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,16384,0.01460444430510203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,12288,0.013071111506885953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,8192,0.007345777418878343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,12288,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,10240,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,7168,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,10240,0.011361777782440186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,8192,0.01036266651418474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,6144,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,7168,0.00998133338159985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,8192,0.009665778113736046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,5120,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,7168,0.009374221993817223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,6144,0.009481777747472128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,4096,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,6144,0.00868088917599784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,3584,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,5120,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,4096,0.00741244438621733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,3072,0.0053386667536364655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,3584,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,3584,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,768,32,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,2560,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,3072,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,65536,0.04708088768853081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,2048,0.0041946665280395085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,3072,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,2560,0.00700444479783376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,1536,0.003552888830502828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,2048,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,2560,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,1024,0.00300177786913183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,1536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,2048,0.006226666685607698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,768,0.0028844444702068963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,512,0.002528888897763358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,1536,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,1024,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,256,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,768,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,1024,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,128,0.0018613334331247543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,768,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,256,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,512,0.0053022222386466126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,64,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,5120,0.008642666869693333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,128,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,256,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,65536,0.012723555167516073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,4096,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,128,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,51200,0.01850044396188524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,64,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,65536,0.046799110041724316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,512,32,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,65536,0.04398577743106418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,51200,0.03734577695528666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,12288,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,51200,0.03542222248183356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,16384,0.01572800013754103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,10240,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,16384,0.014817777607176038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,12288,0.012122666670216454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,8192,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,10240,0.012033777932325998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,10240,0.011015111373530494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,7168,0.005411555369695027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,8192,0.010674667027261523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,8192,0.009703999592198266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,512,512,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,6144,0.005976000179847081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,7168,0.010112889111042023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,512,32,0.0018604445374674266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,5120,0.006562666760550604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,7168,0.00925866680012809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,6144,0.009340444372759925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,4096,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,6144,0.008555555509196388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,3584,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,5120,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,4096,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,16384,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,5120,0.00815822184085846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,3072,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,4096,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,2560,0.005193777796294955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,3584,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,2048,0.004190222256713443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,3072,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,3584,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,2560,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,1536,0.0035457776652442086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,12288,0.013063111239009433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,3072,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,1024,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,2560,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,768,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,2048,0.006246222390068903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,1536,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,512,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,1024,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,1536,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,768,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,1024,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,512,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,768,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,512,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,256,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,64,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,256,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,32,0.0017964444640609955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,128,0.005429333282841577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,65536,0.016872889465755887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,64,0.004970666848950916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,32,0.005008888741334279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,51200,0.014821334017647637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,65536,0.043269332912233144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,65536,0.045997334851158984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,16384,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,51200,0.03736089004410638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,256,2048,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,51200,0.03528177738189697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,12288,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,10240,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,16384,0.015497777197096082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,16384,0.014807111687130399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,8192,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,256,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,12288,0.013117333253224691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,7168,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,10240,0.011769777370823754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,8192,0.01073155552148819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,6144,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,7168,0.010028444230556488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,7168,0.00941955546538035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,256,128,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,5120,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,6144,0.009318222602208456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,4096,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,6144,0.008692444198661381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,5120,0.008978666530715095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,3584,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,4096,0.00812177773978975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,4096,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,3072,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,3584,0.007816000117195977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,3584,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,2560,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,3072,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,3072,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,2048,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,2560,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,12288,0.012104888757069906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,256,128,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,2560,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,1536,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,1024,0.002893333426780171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,2048,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,8192,0.010009778042634329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,1536,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,2048,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,768,0.002552888873550627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,1024,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,1536,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,1024,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,768,0.00600622221827507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,256,0.0019173332386546666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,768,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,512,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,128,0.0018008889423476325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,512,0.004944889081848992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,5120,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,64,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,256,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,32,0.001828444500764211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,256,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,33,128,128,0.00535022219022115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,65536,0.01074488874938753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,51200,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,128,0.004592000196377437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,16384,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,64,0.0046160001721647055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,12288,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,32,0.004595555365085602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,10240,0.00544088871942626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,51200,0.03512888815667894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,16384,0.014450665977266101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,128,10240,0.010963555839326648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,12288,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,7168,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,10240,0.010753778119881948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,128,512,0.002518222149875429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,8192,0.009657777845859528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,5120,0.004558222161398994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,7168,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,4096,0.004512888689835866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,3584,0.004857777721352047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,6144,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,3072,0.004562666846646203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,5120,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,4096,0.007151111132568783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,3584,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,2048,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,3072,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,1536,0.003565333369705412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,1024,0.0028488888508743713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,2560,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,768,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,2048,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,512,0.0024702221983008915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,1024,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,8192,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,256,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,128,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,768,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,64,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,6144,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,512,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,32,0.0018079999006456798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,65536,0.010488000180986194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,256,0.005068444543414646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,51200,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,32,0.004942222187916438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,65536,0.04311022162437439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,12288,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,51200,0.03527466787232293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,10240,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,16384,0.0148062225845125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,8192,0.005532444351249271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,12288,0.012039999994966718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,7168,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,10240,0.01108888867828581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,65536,0.04312800036536323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,1536,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,8192,0.009820444716347588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,5120,0.004568888909286923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,4096,0.004254222330119875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,7168,0.009022222624884712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,3584,0.004592000196377437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,6144,0.008535111116038429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,3072,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,5120,0.008047999607192146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,2560,0.0046302220887608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,64,128,0.005008000052637524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,2048,0.004569777597983678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,3584,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,1536,0.0034871109657817413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,3072,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,16384,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,1024,0.002814222127199173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,768,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,2560,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,512,0.002471999989615546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,2048,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,256,0.0018737777653667664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,1536,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,1024,0.0053084443012873335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,768,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,512,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,64,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,256,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,6144,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,128,0.004576888763242298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,64,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,32,0.004998222407367495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,33,32,4096,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,64,2560,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,51200,0.4918639924791124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,128,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,51200,1.1333529154459636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,16384,0.365847110748291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,33,32,32,0.001828444500764211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,16384,0.16575200027889675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,12288,0.2724960115220812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,12288,0.12319644292195638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,10240,0.22530666987101236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,12288,0.12133688396877712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,10240,0.10422044330173069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,8192,0.18326221572028267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,10240,0.08894311057196723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,12288,0.2077066633436415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,16384,0.2768790986802843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,8192,0.08881244394514297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,7168,0.1621991131040785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,16384,0.1386000050438775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,8192,0.07247200277116564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,10240,0.17470488283369276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,7168,0.074654221534729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,6144,0.1379688845740424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,7168,0.06383822361628215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,5120,0.11696978410085042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,8192,0.13941155539618597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,5120,0.055769774648878306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,6144,0.056624001926845975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,7168,0.12511200375027126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,6144,0.10824977689319187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,4096,0.09390133619308472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,5120,0.05155377917819553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,5120,0.09399377637439305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,51200,0.4533528751797146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,3584,0.0837777786784702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,4096,0.04629955689112345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,4096,0.04012889001104567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,3072,0.07143644491831462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,3584,0.03821422325240241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,4096,0.07932266924116346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,3072,0.04139466749297248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,6144,0.06589244471655951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,3072,0.031344887283113264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,2560,0.05999200211630928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,3584,0.0766142209370931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,2048,0.048342221313052706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,3072,0.06883200009663899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,2560,0.03439822130733066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,2560,0.027423999375767175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,1536,0.03724977705213759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,2560,0.07026755809783936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,2048,0.023504000571038988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,1024,0.026700443691677515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,1536,0.018965333700180054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,1536,0.023050665855407715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,1536,0.04120088948143853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,1024,0.017872888180944655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,768,0.019891555110613506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,1024,0.015134221977657743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,3584,0.04618666569391886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,1024,0.03353510962592231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,768,0.013441777891582914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,512,0.013502221968438892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,768,0.01443377799457974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,256,0.007251555720965068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,51200,0.8817573123508029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,512,0.011858666936556498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,512,0.011940444509188334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,128,0.005306666509972678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,512,0.02409155501259698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,256,0.009331555830107795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,2048,0.03015111221207513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,256,0.023205333285861548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,64,0.004279110994603899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,65536,32,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,128,0.010448888772063784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,128,0.008007999923494127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,128,0.015484443969196744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,64,0.01071111112833023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,2048,0.04980088935958015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,65536,32,0.012473777764373355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,768,0.028432889117134943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,65536,0.5237760014004177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,51200,0.3951884375678168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,256,0.010303999814722273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,65536,1.0959190792507596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,51200,0.8452746603224012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,16384,0.2772871123419868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,16384,0.1326542165544298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,12288,0.20991999573177764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,65536,0.4376844565073649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,12288,0.09945778052012126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,16384,0.11041777663760716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,51200,0.3574639956156413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,10240,0.1726133293575711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,12288,0.083588441212972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,10240,0.08449421988593207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,10240,0.07022311290105183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,8192,0.14128888977898493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,8192,0.07214488585789998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,12288,0.16996888319651285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,16384,0.22351378864712187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,7168,0.12315200434790717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,7168,0.06075555748409695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,8192,0.06169333060582479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,8192,0.11734844578637017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,6144,0.10583021905687119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,7168,0.05259644322925144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,6144,0.05350222190221151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,5120,0.08975999885135227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,7168,0.1046560009320577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,6144,0.04612000121010674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,4096,0.07276266813278198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,6144,0.09273511171340942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,5120,0.045208000474505954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,10240,0.14195999834272596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,5120,0.04233333468437195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,4096,0.03762755460209317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,3584,0.06378755304548475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,4096,0.03297777639495002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,3072,0.05558222532272339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,3584,0.03337511089113023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,4096,0.06741511159472995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,3584,0.06173155705134074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,3072,0.025949334104855854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,2560,0.04693422383732266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,3072,0.03308355477121141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,3072,0.055883553293016225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,2560,0.023238221804300945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,2048,0.03764977720048692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,2560,0.02959377898110284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,51200,0.689175976647271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,2560,0.058055996894836426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,1536,0.029719998439153034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,2048,0.01942222151491377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,2048,0.027063111464182537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,1024,0.020958221620983545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,2048,0.0426862206723955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,5120,0.08097244633568658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,1536,0.016138666205936007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,768,0.015477332803938123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,1536,0.019084443648656208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,1024,0.014834667245546976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,1024,0.015559110376569959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,3584,0.03972355524698893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,512,0.010800888968838586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,65536,0.895200941297743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,1024,0.0288426677385966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,1536,0.03870666689342923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,256,0.006467555546098285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,768,0.0123857781291008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,768,0.024698666400379602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,128,0.003882666842805015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,512,0.011622221933470832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,256,0.009324444664849175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,256,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,512,0.02103022237618764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,64,0.0028239999794297745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,51200,32,0.00314666661951277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,51200,128,0.009371555513805812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,128,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,128,0.01479022204875946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,64,0.008985777695973715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,32,0.01076533314254549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,65536,0.3566888968149821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,65536,0.18115644984775117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,51200,0.25786132282680935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,65536,0.16040533118777806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,768,0.012668444050682915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,51200,0.1327306694454617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,16384,0.07913511329227023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,51200,512,0.010661332971519895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,12288,0.06062489085727268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,16384,0.0436062216758728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,16384,0.048900445302327476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,51200,256,0.01974933346112569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,10240,0.05054488778114319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,12288,0.03286222285694546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,12288,0.03681244452794393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,12288,0.04344355397754245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,10240,0.0321999986966451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,8192,0.042698668109046094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,10240,0.03894844320085313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,51200,0.13998399840460882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,8192,0.027667555544111464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,7168,0.03640000025431315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,8192,0.025906667113304138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,8192,0.03513866662979126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,65536,0.1826159954071045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,6144,0.03172355559137132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,7168,0.022076444493399724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,7168,0.024290665984153748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,51200,0.1292977730433146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,6144,0.02184266679816776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,7168,0.03047200043996175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,5120,0.02721333338154687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,6144,0.027863999207814533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,4096,0.023200889428456623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,16384,0.056405332353379994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,5120,0.018623110320832994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,5120,0.018903111418088276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,3584,0.019882667395803664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,5120,0.025128000312381323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,4096,0.015384889311260648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,4096,0.016082665986484952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,3072,0.01755644381046295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,4096,0.021984888447655573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,3584,0.014660444524553088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,3584,0.017542221479945712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,2560,0.01481599940194024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,3072,0.015282667345470853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,3072,0.013288000391589271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,3584,0.020223110914230347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,2048,0.01204622205760744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,2560,0.01219733307758967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,3072,0.018582221534517076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,2560,0.017626666360431247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,2048,0.011053333679835001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,1536,0.010120000276300643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,6144,0.019860444797409903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,2048,0.016536888149049547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,1536,0.009429333110650381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,1536,0.010395555860466426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,1536,0.01447111037042406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,10240,0.02934044599533081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,768,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,1024,0.008623999853928884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,1024,0.007903111477692923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,1024,0.01311644415060679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,512,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,768,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,768,0.007000000112586551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,768,0.012175111307038201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,256,0.0028328889360030494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,512,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,512,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,512,0.009993777506881291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,128,0.0024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,256,0.0063075555695427795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,2048,0.01257599973016315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,256,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,64,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,256,0.008984888593355814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,32,0.0021973333011070886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,16384,1024,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,128,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,64,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,32,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,65536,0.2547644509209527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,65536,0.14984622266557482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,51200,0.20212711228264701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,65536,0.13296177652147081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,51200,0.11461244689093696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,16384,0.07144355773925781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,2560,0.014523555835088095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,51200,0.1158871120876736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,16384,0.042118221521377563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,12288,0.0561715563138326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,128,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,16384,0.04695644312434726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,12288,0.03234133455488417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,10240,0.045906666252348155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,12288,0.029976000388463337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,12288,0.03775733378198411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,65536,0.14698933230506048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,8192,0.03828888800409105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,51200,0.11475111378563775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,10240,0.026050666968027752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,10240,0.028006222512986925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,7168,0.03347288899951511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,10240,0.03377244538731045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,8192,0.021725333399242822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,6144,0.029457777738571167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,8192,0.02459200057718489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,7168,0.020067556036843192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,8192,0.0299928883711497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,5120,0.021756443712446425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,7168,0.02676800058947669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,6144,0.01796444422668881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,6144,0.01891377733813392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,4096,0.018737778067588806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,16384,0.0377128885851966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,5120,0.016765332884258695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,5120,0.02165422174665663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,16384,128,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,3584,0.015786666009161208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,4096,0.014317333698272705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,4096,0.014436443646748861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,3072,0.014074666632546319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,4096,0.019707555572191875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,3584,0.016216889023780823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,3584,0.0184551113181644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,3072,0.014142221874660917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,3072,0.012499555945396423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,2560,0.012913777596420713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,3072,0.017251556118329365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,7168,0.021380444367726643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,2560,0.011401777466138205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,2048,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,2560,0.013134222063753339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,2560,0.01647555496957567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,1536,0.008625778059164682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,2048,0.009932444327407414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,6144,0.02477866742346022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,5120,0.016553777787420485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,1024,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,2048,0.015593777100245157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,1536,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,1536,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,768,0.0058577776783042485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,1536,0.013824889229403602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,1024,0.008019555773999956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,3584,0.013406222065289816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,512,0.004087111188305749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,1024,0.011723555624485016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,768,0.00720355576939053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,768,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,256,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,768,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,512,0.00665511108107037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,512,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,128,0.002493333278430833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,512,0.010315555665228102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,256,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,256,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,64,0.0022515555222829184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,256,0.009038222332795462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,128,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,12288,32,0.002263111165828175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,128,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,128,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,64,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,2048,0.011507555842399597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,32,0.006543111056089401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,51200,0.18243911531236437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,65536,0.1436800029542711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,12288,1024,0.00741244438621733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,65536,0.12410844696892633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,51200,0.10908710956573486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,16384,0.0607280002699958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,51200,0.10063199864493476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,16384,0.03563377923435635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,12288,0.04664533336957296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,16384,0.04041422075695462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,16384,0.043509334325790405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,12288,0.028524445162879095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,10240,0.039657778210110135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,12288,0.03081511126624213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,12288,0.036171555519104004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,8192,0.0339662233988444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,65536,0.13412355052100286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,51200,0.10784800185097589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,10240,0.026926222774717543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,65536,0.22903998692830405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,7168,0.027502222193611994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,8192,0.021470222208235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,10240,0.03181155522664388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,8192,0.023208000593715247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,7168,0.019188443819681805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,7168,0.020145777199003432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,5120,0.01954844428433312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,6144,0.01754133403301239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,7168,0.025952888859642878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,6144,0.023756444454193115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,4096,0.016601777738995023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,5120,0.016536000702116225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,5120,0.015827554795477126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,5120,0.020999999509917364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,3584,0.01500266624821557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,4096,0.01402844488620758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,4096,0.01904622217019399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,3584,0.013758222262064615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,3584,0.013411555853154926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,3072,0.013376889129479727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,10240,0.024098666177855596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,3584,0.018648000227080453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,2560,0.010692444112565784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,6144,0.023052444060643513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,3072,0.01180266671710544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,8192,0.02789155642191569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,3072,0.012774222426944308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,2048,0.009773333039548662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,2560,0.01035822182893753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,6144,0.018191999859280057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,1536,0.008036444584528605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,2048,0.008993777963850234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,2048,0.010517333116796283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,2048,0.014866666661368476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,1024,0.0063386667105886675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,1536,0.008100444244013892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,1536,0.008431110945012834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,4096,0.013968000809351603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,1536,0.013734222286277346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,768,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,1024,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,1024,0.011208000282446543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,768,0.006573333508438534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,768,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,512,0.003958222352796131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,768,0.010032888915803697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,512,0.006646222124497096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,512,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,256,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,512,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,2560,0.01645866698688931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,128,0.0025271111064487034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,256,0.006153777655627992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,256,0.005387555393907759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,256,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,64,0.002191999927163124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,128,0.006535999890830781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,128,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,10240,32,0.0022968887868854734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,128,0.007326222128338284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,1024,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,64,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,65536,0.17259644137488472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,32,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,51200,0.12525510787963867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,65536,0.09178133143319024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,65536,0.07033689154518975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,51200,0.0707448919614156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,16384,0.04583911101023356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,10240,2560,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,51200,0.058992001745435924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,12288,0.03488622109095255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,16384,0.0229777776532703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,16384,0.028022223048739966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,16384,0.042711112234327525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,65536,0.1324862241744995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,12288,0.018562666243977018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,10240,0.026956443985303242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,12288,0.021877333521842957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,8192,0.022798221972253587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,51200,0.10646577676137288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,12288,0.03571733170085483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,10240,0.016498666670587327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,10240,0.019277332557572257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,7168,0.019865777757432725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,8192,0.015220445063379077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,10240,0.03137333194414774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,8192,0.016524444023768108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,6144,0.01749955614407857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,8192,0.027052443888452318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,7168,0.01369244439734353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,7168,0.014665777484575907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,7168,0.025009777810838487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,6144,0.013401777380042605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,6144,0.0125591109196345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,4096,0.012721777790122561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,6144,0.022836445106400385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,5120,0.01204711116022534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,5120,0.01199288914601008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,3584,0.011007111105653973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,4096,0.010692444112565784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,4096,0.010737777584128909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,3072,0.009992889232105678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,4096,0.01864622202184465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,3584,0.011357333097192975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,3584,0.010111110905806223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,2560,0.00866844422287411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,3584,0.01753777762254079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,3072,0.009008888569143083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,3072,0.010626667075686984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,2048,0.007426666716734569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,3072,0.016945777667893302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,2560,0.008271111382378472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,2560,0.00977955593003167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,2560,0.015535111228624979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,2048,0.008757332960764566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,5120,0.014935111006100973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,1536,0.007327111230956183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,3072,0.017190222938855488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,1024,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,1536,0.006984000404675801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,1536,0.012654222548007965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,768,0.003621333175235324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,1024,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,5120,0.020505777663654752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,1024,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,1024,0.011010666688283285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,512,0.0029226665695508323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,768,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,768,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,256,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,768,0.010688888529936472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,512,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,512,0.005623111294375525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,128,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,512,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,256,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,256,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,1536,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,256,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,2048,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,128,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,128,0.004942222187916438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,32,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,2048,0.014537778165605334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,128,0.0070088886552386815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,64,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,65536,0.1521057817671034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,8192,32,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,51200,0.12039111720191108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,65536,0.06660444206661649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,16384,0.04236800140804715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,51200,0.06447911262512207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,51200,0.054141332705815635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,16384,0.025808889004919264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,12288,0.03327644533581204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,16384,0.02167466613981459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,65536,0.13224889172448054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,51200,0.1061191095246209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,16384,0.042542222473356456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,10240,0.028169777658250596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,8192,64,0.001995555642578337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,12288,0.01777777738041348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,12288,0.020040889581044514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,8192,0.023870221442646448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,10240,0.015815110670195687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,10240,0.01773066653145684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,12288,0.03519999980926514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,7168,0.018193778064515855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,10240,0.030672887961069744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,8192,0.014076444837782117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,8192,0.015180443723996481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,6144,0.01590844492117564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,65536,0.08394222127066718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,8192,0.02623822291692098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,7168,0.013749333719412485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,5120,0.0136817776494556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,7168,0.024357333779335022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,6144,0.012157333393891653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,6144,0.012763555679056378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,4096,0.011977777712874942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,6144,0.022092445029152766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,5120,0.011509333219793109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,3584,0.010270222193664974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,5120,0.020466667082574632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,4096,0.00962933318482505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,4096,0.010194666683673859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,3072,0.009210666848553551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,4096,0.018576888574494254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,3584,0.008938666847017076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,3584,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,2560,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,3584,0.017311111092567444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,3072,0.00831733312871721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,3072,0.010082667072614035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,2560,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,3072,0.01679733395576477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,2560,0.009296000003814697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,2560,0.015377778146002026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,2048,0.007805333369308048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,2048,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,7168,0.012980444563759698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,1024,0.004544888933499654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,2048,0.014190221826235453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,1536,0.006917333437336816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,1536,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,768,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,5120,0.011374221907721626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,1024,0.006191111273235745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,1024,0.010666666759385003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,512,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,768,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,768,0.005988444305128521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,768,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,512,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,512,0.005495999836259418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,2048,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,256,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,256,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,128,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,256,0.008972444468074376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,128,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,128,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,64,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,32,0.002116444417172008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,128,0.007277333074145847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,1536,0.012079999678664737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,65536,0.12285066975487603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,64,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,7168,32,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,51200,0.09689866834216648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,65536,0.07686577902899848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,256,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,65536,0.06411200099521212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,16384,0.03612711032231649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,512,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,51200,0.05801688962512546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,51200,0.05244000090493096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,7168,1536,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,16384,0.02387822171052297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,65536,0.1321822272406684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,16384,0.020890666378868952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,16384,0.04254133171505398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,51200,0.10611110925674438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,10240,0.024683554967244465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,12288,0.017257778180970084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,12288,0.018418666389253404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,1024,0.0064488889442549805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,12288,0.034309334225124784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,10240,0.01517688896920946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,10240,0.016167999969588388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,7168,0.018232888645595975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,10240,0.030142221185896132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,8192,0.013507555756303998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,8192,0.014344000154071383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,6144,0.016167110866970487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,8192,0.0258942229880227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,7168,0.012635555532243518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,7168,0.013005333642164866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,5120,0.014142221874660917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,7168,0.023732443650563557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,6144,0.011715555356608497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,6144,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,4096,0.010988444089889526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,6144,0.02197422252760993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,5120,0.010625777973069085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,5120,0.010878221856223213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,12288,0.02858755654758877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,3584,0.009422221945391761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,5120,0.02000266644689772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,4096,0.008973333570692275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,4096,0.009983110758993361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,3584,0.008543111383914948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,4096,0.017972444494565327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,8192,0.02055111030737559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,3584,0.017302221722073026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,2560,0.007541333635648091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,3072,0.009448889229032729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,3072,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,2048,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,3072,0.01615111033121745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,2560,0.007583111524581909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,2560,0.008848889006508721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,2560,0.014500444134076437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,1536,0.006014222072230444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,2048,0.007234666910436418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,2048,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,1024,0.004242666479614046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,2048,0.013421333498424955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,1536,0.006783111227883234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,1536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,768,0.00350577798154619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,1536,0.012125333150227865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,1024,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,1024,0.010392888552612728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,768,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,512,0.002879111096262932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,768,0.005884444548024072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,768,0.00998488896422916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,3584,0.010372444159454769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,256,0.00250133333934678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,512,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,128,0.002128888956374592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,512,0.005762666463851929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,512,0.009326222042242685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,256,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,256,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,256,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,128,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,32,0.00198399999903308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,128,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,65536,0.10898844401041667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,64,0.005480000128348668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,6144,32,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,51200,0.08669422070185344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,65536,0.06999378071890937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,1024,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,65536,0.06165511078304715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,16384,0.031416889694001936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,3072,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,51200,0.049984001451068454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,51200,0.052686221069759794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,12288,0.025389333566029865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,16384,0.021423111359278362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,65536,0.13218400213453504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,16384,0.02017600006527371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,10240,0.021686222818162706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,16384,0.04208799865510729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,12288,0.01677688956260681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,12288,0.016864889197879367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,128,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,8192,0.018917333748605516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,12288,0.034128000338872276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,10240,0.014811555544535318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,10240,0.014914666612943014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,7168,0.016538666354285348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,10240,0.02944799926545885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,8192,0.013167999684810638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,8192,0.013925333817799887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,6144,0.014466666513019137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,8192,0.025795555777019922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,7168,0.012321777641773224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,7168,0.012273777690198688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,5120,0.012855110896958245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,7168,0.023560000790490046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,6144,0.011362666885058085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,6144,0.011706666813956367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,4096,0.009675555759006077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,5120,0.00981600003110038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,5120,0.010315555665228102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,5120,0.019985778464211356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,3584,0.008488000267081791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,4096,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,4096,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,51200,0.10573422246509129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,3072,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,4096,0.018235555953449674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,6144,64,0.002020444514022933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,3584,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,2560,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,3072,0.007727999654081132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,3584,0.017089777522616915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,3072,0.009010666774378883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,2048,0.00610133343272739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,3072,0.015839111473825242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,2560,0.00758755538198683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,2560,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,1536,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,2048,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,2560,0.014952888091405233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,2048,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,1024,0.0041164445380369825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,2048,0.01277955538696713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,1536,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,6144,0.021573333276642695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,768,0.003504000190231535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,1024,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,1536,0.012033777932325998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,1024,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,512,0.0032106666929192017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,1024,0.010320888625250923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,768,0.006067555397748947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,768,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,256,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,512,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,768,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,3584,0.009685333404276106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,128,0.002191999927163124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,512,0.009597332941161262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,256,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,256,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,64,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,128,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,256,0.009305777649084726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,128,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,128,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,65536,0.0870328876707289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,64,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,32,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,51200,0.0695271094640096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,65536,0.061843554178873696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,65536,0.059622221522861056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,16384,0.026312000221676294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,51200,0.04827022221353319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,51200,0.048022223843468555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,65536,0.13183289104037815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,12288,0.019779556327395968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,51200,0.10559111171298557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,5120,512,0.0053004444473319584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,16384,0.019520888725916546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,16384,0.01942044496536255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,10240,0.017235555582576327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,16384,0.041722665230433144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,12288,0.01605333387851715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,12288,0.015562666787041558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,8192,0.01482222146458096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,10240,0.013932444983058505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,10240,0.014452444182501899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,5120,32,0.0019315555691719055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,7168,0.013186666700575085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,10240,0.029543999168607924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,8192,0.012644444074895648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,8192,0.012777778009573618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,6144,0.012082666986518435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,8192,0.025397333833906386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,1536,0.006637333167923822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,7168,0.011704888608720569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,5120,0.011580444044537015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,7168,0.02368444369898902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,6144,0.010669333239396414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,4096,0.010100444157918295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,6144,0.021702221698231165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,5120,0.00964355551534229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,5120,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,3584,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,5120,0.019875556230545044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,4096,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,4096,0.008611555728647444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,4096,0.01758666667673323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,3584,0.008908444808589088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,3584,0.008046222229798635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,3584,0.01682133310370975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,12288,0.033187554942237005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,2560,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,3072,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,3072,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,3072,0.015793777174419828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,2048,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,7168,0.011717333561844297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,2560,0.007328888608349695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,2560,0.006907555378145642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,1536,0.003943110919660992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,6144,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,2048,0.007002666592597961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,2048,0.006561777657932705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,1024,0.003206222214632564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,2048,0.012744888663291931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,1536,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,1536,0.011390222443474663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,768,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,1024,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,1024,0.01109688894616233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,3072,0.008590222232871586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,512,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,768,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,768,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,768,0.010059555371602377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,256,0.0022408889813555614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,512,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,512,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,256,0.004957333207130432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,128,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,2560,0.013795555465751223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,256,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,256,0.0087333329849773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,64,0.0019324444648292328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,1536,0.006637333167923822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,4096,32,0.0019662222928471034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,128,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,128,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,64,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,1024,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,4096,32,0.00543200017677413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,51200,0.059283554553985596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,65536,0.058693336115943066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,65536,0.05952622493108114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,16384,0.023144889209005568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,512,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,51200,0.04695999953481886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,65536,0.13193066914876303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,51200,0.046747555335362755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,12288,0.018625777628686693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,16384,0.0192266669538286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,51200,0.10594311025407578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,16384,0.018630221486091614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,10240,0.015812445018026564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,16384,0.04124177826775445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,12288,0.0158924443854226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,8192,0.013749333719412485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,128,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,12288,0.032947556840048894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,10240,0.014462222655614218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,7168,0.012424888710180918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,65536,0.07374399900436401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,8192,0.012235555383894177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,10240,0.02936089038848877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,8192,0.012372444073359171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,6144,0.01148799972401725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,8192,0.02537244392765893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,7168,0.011219555305110084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,5120,0.011391110718250275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,6144,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,7168,0.023360000716315374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,6144,0.010451555252075195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,4096,0.009136888715955947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,6144,0.02197422252760993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,5120,0.00943022221326828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,5120,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,3584,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,4096,0.008621333373917473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,5120,0.0195777780479855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,4096,0.017849778135617573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,12288,0.015104000767072042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,3584,0.008024888734022776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,3584,0.01591822173860338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,10240,0.013578666581047906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,2560,0.006607999818192587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,3072,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,3072,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,3072,0.015480000111791821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,7168,0.011114666859308878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,2048,0.005076444397370021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,2560,0.007423111134105259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,2560,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,1536,0.0039048890272776284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,2560,0.013806222213639153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,2048,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,2048,0.0063324446479479475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,2048,0.012776888906955719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,1536,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,1536,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,4096,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,1536,0.011724444727102915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,768,0.0028853333658642242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,1024,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,1024,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,3584,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,1024,0.010348444183667501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,512,0.0025617778301239014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,768,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,768,0.005524444497293896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,256,0.002320000073975987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,768,0.010032888915803697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,512,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,512,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,128,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,512,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,256,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,256,0.005048888838953442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,64,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,256,0.009016888837019602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,128,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,32,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,128,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,65536,0.06506399975882636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,1024,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,64,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3584,32,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,51200,0.04993066522810194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,65536,0.05711288584603203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,65536,0.057499554422166615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,16384,0.019725332657496136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,51200,0.0461715559164683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,51200,0.044884443283081055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,16384,0.017882666654056974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,51200,0.10590755277209812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,16384,0.01905155513021681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,12288,0.016588444511095684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,16384,0.0406640006436242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,12288,0.015830222103330824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,10240,0.013923555612564087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,12288,0.014733332726690503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,12288,0.03279733326699998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,8192,0.012068444656001197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,10240,0.014025777578353882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,10240,0.013208889298968844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,128,0.007340444458855524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,7168,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,10240,0.029168887270821467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,8192,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,8192,0.012336889074908363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,6144,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3584,3072,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,8192,0.02551733288500044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,7168,0.011064888702498542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,5120,0.009704888694816166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,7168,0.02334844403796726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,65536,0.131750226020813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,6144,0.010231110784742568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,4096,0.00868711123863856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,5120,0.008967999782827165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,5120,0.009008888569143083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,3584,0.00815466625822915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,4096,0.00831911133395301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,5120,0.01907911068863339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,4096,0.01682488951418135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,3072,0.0074720001882976955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,3584,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,3584,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,3584,0.015495111544926962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,2560,0.006507555643717448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,3072,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,3072,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,3072,0.014533332652515836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,2048,0.0046364445653226645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,2560,0.007332444190979004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,7168,0.010777778095669217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,2560,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,1536,0.0041857775714662345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,2560,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,2048,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,2048,0.0069875551594628235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,6144,0.01018488903840383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,1024,0.0032239999208185407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,2048,0.012452444268597497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,1536,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,6144,0.021908443835046556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,768,0.0028977776981062363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,1536,0.011400889191362592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,1024,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,512,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,1024,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,1024,0.010729778144094678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,768,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,768,0.0052800000541739995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,768,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,512,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,512,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,128,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,512,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,256,0.005530666559934616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,256,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,64,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,256,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,128,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,3072,32,0.0018986666368113623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,128,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,65536,0.05758844481574165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,128,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,64,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,32,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,51200,0.044891556104024254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,65536,0.0550124446551005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,65536,0.05665333403481377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,1536,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,65536,0.13224177890353733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,51200,0.04288088944223192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,51200,0.10526488886939155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,16384,0.01772444446881612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,16384,0.018633777896563213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,12288,0.014122666584120857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,12288,0.01512888901763492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,16384,0.04076888826158311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,12288,0.014514666464593677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,12288,0.032856888241238065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,10240,0.01276266657643848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,10240,0.013136000268989138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,10240,0.029109332296583388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,16384,0.018385777870814007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,8192,0.011635555161370171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,7168,0.010372444159454769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,8192,0.02528088953759935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,51200,0.0454675555229187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,8192,0.011375111010339526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,6144,0.00960533320903778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,7168,0.010384889112578498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,7168,0.010640000303586325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,7168,0.023005333211686876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,6144,0.009655111365848118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,5120,0.008777778181764815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,6144,0.00960799968904919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,10240,0.01255644443962309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,4096,0.008046222229798635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,6144,0.021233777205149334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,5120,0.008790222307046255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,3584,0.007518221934636434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,5120,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,4096,0.0083155557513237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,4096,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,5120,0.019000000423855253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,3072,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,4096,0.016529776983790927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,3584,0.007954667011896769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,3584,0.007624000310897827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,3072,0.007383999725182851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,3584,0.015857777661747403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,3072,0.014667555689811707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,2048,0.004398222184843487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,3072,4096,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,2560,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,2560,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,1536,0.003937777959638172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,2048,0.006991110742092133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,2560,0.013414222333166333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,2048,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,1024,0.003173333282272021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,2048,0.012423999607563019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,1536,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,1024,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,768,0.002847111059559716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,1024,0.010372444159454769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,8192,0.011295999917719098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,768,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,768,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,2560,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,3072,0.006912888752089606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,512,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,512,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,256,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,512,0.009670221971140968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,256,0.005409777578380372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,128,0.0020995556066433587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,256,0.005212444398138259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,1536,0.0063982220987478895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,256,0.008683555656009251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,128,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,1536,0.011695110963450538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,32,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,1024,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,128,0.007595555649863348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,65536,0.046557333734300405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,64,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,51200,0.03643555442492167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,32,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,65536,0.055493334929148354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,65536,0.052424887816111244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,768,0.01001422190003925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,16384,0.01511111193233066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,51200,0.041448887851503156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,51200,0.044680890109803945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,65536,0.13145955403645834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,12288,0.012078222301271228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,51200,0.10558578040864731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,16384,0.01853155593077342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,16384,0.017095999585257638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,10240,0.010831111007266574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,12288,0.013853333062595792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,12288,0.03297066688537598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,10240,0.012433778080675336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2560,128,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,8192,0.011852444046073489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,10240,0.013080889152155982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,10240,0.028692444165547688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,7168,0.010752889017264048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,8192,0.011081777513027191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,8192,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2560,512,0.0025048889219760895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,8192,0.02539288832081689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,6144,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,7168,0.01033333357837465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,7168,0.010095111197895473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,5120,0.009116444322797988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,7168,0.02365866634580824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,6144,0.009331555830107795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,6144,0.009179555707507664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,4096,0.008606221940782335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,6144,0.02163644466135237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,5120,0.008624888956546783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,5120,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,3584,0.00776444458299213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,16384,0.04036533170276218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,4096,0.008007110820876228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,5120,0.01870844430393643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,4096,0.0075235557225015424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,3072,0.0057760001056724125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,3584,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,4096,0.016550223032633465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,2560,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,3584,0.015456888410780164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,3072,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,3072,0.01478488908873664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,2048,0.004304888761705822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,2560,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,2560,0.00701155596309238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,2560,0.013439111411571503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,1536,0.0038364442686239877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,2048,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,2048,0.012568000290128918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,1024,0.0031973332580592898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,1536,0.006308444258239534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,1536,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,1536,0.011758222348160215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,12288,0.014707555373509726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,1024,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,1024,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,1024,0.01034577770365609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,512,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,768,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,3584,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,768,0.0053342220683892565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,768,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,256,0.0021395554973019496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,3072,0.007320000065697565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,512,0.005707555760939916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,128,0.0019466665883858998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,512,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,256,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,512,0.010019555687904358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,64,0.001826666709449556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,256,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,32,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,256,0.009411555197503831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,128,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,128,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,65536,0.035712000396516584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,128,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,64,0.005070222334729301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,2048,32,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,65536,0.0549759997261895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,65536,0.050352000527911715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,16384,0.012488889197508493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,2048,768,0.0028320000403457214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,65536,0.13139022721184626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,51200,0.04015644391377767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,51200,0.04435111085573832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,12288,0.010772444307804108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,51200,0.10493866602579753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,16384,0.01751377847459581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,16384,0.016608888904253643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,10240,0.01514400045077006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,16384,0.04049688908788893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,12288,0.013386666774749756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,8192,0.013349333571063148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,12288,0.032906668053732976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,10240,0.011935110721323224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,10240,0.012228444218635559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,10240,0.02884622083769904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,7168,0.012410666379663678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,8192,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,8192,0.010384889112578498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,8192,0.025224000215530396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,6144,0.011000889043013254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,7168,0.010031999813185798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,7168,0.009707555174827576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,7168,0.022288888692855835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,6144,0.009511999785900116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,5120,0.010013333625263637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,6144,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,6144,0.020120888948440552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,4096,0.00889511075284746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,5120,0.009004444711738164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,5120,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,5120,0.01817244456874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,12288,0.013603555659453074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,3584,0.008958222137557136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,4096,0.007369777394665613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,4096,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,3072,0.005747555444637935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,4096,0.016519111063745286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,3584,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,3584,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,2048,0.006831999868154526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,2560,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,3072,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,3584,0.015455999308162265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,3072,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,2048,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,3072,0.014816888504558139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,51200,0.02792888879776001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,2560,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,1536,0.0037875556283526947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,2560,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,2048,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,2560,0.013456000222100152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,2048,0.006302222195598815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,1024,0.003175111073586676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,2048,0.01242844429281023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,1536,0.00664533343580034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,1536,0.0057760001056724125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,1024,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,1536,0.01107911103301578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,512,0.002463111033042272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,1024,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,1024,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,768,0.0052826665341854095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,256,0.002191999927163124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,512,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,512,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,512,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,128,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,256,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,256,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,256,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,64,0.0018151110659042995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,128,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,32,0.0018648888087934917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,128,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,65536,0.025430222352345783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,64,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1536,32,0.004960000101062986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1536,768,0.0028071111689011254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,51200,0.020636444290479023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,65536,0.04849955439567566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,65536,0.054280890358818897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,16384,0.011443555355072021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,768,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,51200,0.04383644461631775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,768,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,12288,0.009675555759006077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,51200,0.1045831110742357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,16384,0.016141333513789706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,16384,0.015483554866578845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,10240,0.0087013335691558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,16384,0.04056533177693685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,12288,0.013160000244776407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,12288,0.01295644458797243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,8192,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,128,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,12288,0.031926221317715116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,10240,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,7168,0.00740799970097012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,10240,0.028679112593332928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,8192,0.010738666686746808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,8192,0.010025777750545079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,6144,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,8192,0.024656888511445787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,7168,0.009416888985368941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,7168,0.010023111270533668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,5120,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,7168,0.022312000393867493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,6144,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,6144,0.008683555656009251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,65536,0.13146311706966823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,4096,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,51200,0.03889244463708665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,6144,0.020282665888468426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,5120,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,3584,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,4096,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,4096,0.0073902226156658595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,4096,0.01605688863330417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,3072,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,3584,0.007138667007287343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,3584,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,3584,0.015173332558737861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,2560,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,3072,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,3072,0.014150222142537435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,2560,0.00656355544924736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,2048,0.005094222310516569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,2560,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,2560,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,2048,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,1536,0.004407999830113517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,5120,0.008011555506123437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,2048,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,5120,0.0184897780418396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,1024,0.0032355555643637977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,1536,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,1536,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,1536,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,768,0.003136000078585413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,1024,0.005995555470387141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,1024,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,10240,0.012053333222866058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,3072,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,512,0.0025235555238193935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,1024,0.01036622209681405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,768,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,768,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,512,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,768,0.009653333160612319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,512,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,128,0.0021520000365045336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,512,0.010012444522645738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,256,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,256,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,64,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,256,0.00865955568022198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,32,0.0021537778278191886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,2048,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,65536,0.0203813330994712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,64,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,32,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,51200,0.031192888816197712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,65536,0.047248890002568565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,65536,0.05428444345792135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,65536,0.13087022304534912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,51200,0.03834933373663161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,51200,0.04356888930002848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,16384,0.013448889056841532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,1024,256,0.002208000048995018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,12288,0.0088764445649253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,51200,0.1039866672621833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,16384,0.015832889411184523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,16384,0.015166223049163818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,10240,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,16384,0.04008000095685323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,12288,0.01310400002532535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,128,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,8192,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,1024,128,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,10240,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,128,0.007558222446176741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,10240,0.011076444553004371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,7168,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,10240,0.02829066581196255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,8192,0.009763555394278632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,6144,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,8192,0.023349333140585158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,7168,0.010144888526863521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,7168,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,5120,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,7168,0.02200711104604933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,6144,0.009370666411187913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,6144,0.00868711123863856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,4096,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,6144,0.019906666543748643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,5120,0.008897778060701158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,5120,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,3584,0.005705777969625261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,5120,0.018588445252842374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,4096,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,4096,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,3072,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,4096,0.016145777371194627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,3584,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,3584,0.007171555525726742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,2560,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,12288,0.031503111124038696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,3584,0.01551555593808492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,3072,0.007592889169851939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,3072,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,2048,0.004435555388530095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,2560,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,3072,0.014160888062583076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,8192,0.0107004443804423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,2560,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,1536,0.004037333445416557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,2560,0.013286222186353473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,2048,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,2048,0.006177777631415262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,1024,0.0031715554909573663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,2048,0.012729778057999082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,1536,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,1536,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,768,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,1536,0.011424000064531961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,1024,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,512,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,1024,0.011028444601429833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,768,0.005684444473849402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,768,0.005419555637571547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,256,0.002163555473089218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,768,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,512,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,512,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,128,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,12288,0.012425777812798819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,256,0.005294222384691238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,256,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,64,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,256,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,128,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,128,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,768,32,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,65536,0.027161777019500732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,128,0.007463999920421177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,32,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,65536,0.05373511049482557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,65536,0.04687644375695122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,65536,0.1306462287902832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,51200,0.03872977693875631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,16384,0.008707555631796518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,768,1024,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,51200,0.03767555620935228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,12288,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,51200,0.10392533408270942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,16384,0.015848888291252982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,16384,0.014537778165605334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,10240,0.006393777827421824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,12288,0.013045333325862885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,12288,0.012159999873903064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,512,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,8192,0.005642666584915585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,12288,0.03160533308982849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,10240,0.011732444167137146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,10240,0.011018666956159802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,7168,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,10240,0.027465777264700994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,8192,0.0107342220014996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,8192,0.009936889012654623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,6144,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,8192,0.02386933399571313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,51200,0.021977777282396953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,7168,0.010031999813185798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,7168,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,5120,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,7168,0.021584888299306233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,4096,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,5120,0.008683555656009251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,6144,0.00870488915178511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,5120,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,3584,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,5120,0.01789955629242791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,4096,0.008036444584528605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,4096,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,3072,0.005538666827811136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,4096,0.016189333465364244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,3584,0.007663110891977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,16384,0.03958844476275974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,3584,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,3584,0.01512977812025282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,3072,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,3072,0.0069146665434042616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,2048,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,3072,0.014135110709402295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,2560,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,2560,0.006341333190600078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,1536,0.003546666767862108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,2560,0.013414222333166333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,2048,0.006806222101052602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,2048,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,1024,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,2048,0.012239110966523489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,1536,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,1536,0.005631999837027655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,6144,0.00925422211488088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,1536,0.011083555718262991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,1024,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,1024,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,6144,0.02027555637889438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,512,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,768,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,768,0.009883555273214975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,512,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,2560,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,512,0.009367110828558603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,256,0.005048888838953442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,256,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,128,0.0019102222803566193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,256,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,768,0.0028026666906144884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,128,0.004549333204825719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,64,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,128,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,128,0.005321777943107817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,32,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,65536,0.01614755557643043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,1024,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,64,0.004588444613748127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,512,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,512,256,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,65536,0.04597244328922696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,65536,0.046563555796941124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,65536,0.1296737723880344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,51200,0.037626667155159846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,16384,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,51200,0.03597066799799601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,51200,0.10328266355726455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,16384,0.01584533353646596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,12288,0.005380444642570283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,16384,0.014515555567211576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,10240,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,16384,0.0387644436624315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,12288,0.013023110727469126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,12288,0.012366222010718452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,8192,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,12288,0.03152889013290405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,10240,0.011014222270912595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,768,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,10240,0.027457776996824477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,7168,0.005323555734422471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,8192,0.010435555544164447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,8192,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,512,32,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,6144,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,51200,0.013790222505728403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,8192,0.023766222927305434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,7168,0.009872888525327047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,5120,0.004951111144489712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,7168,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,7168,0.02162666618824005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,6144,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,6144,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,4096,0.004563555535342959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,5120,0.008534222013420528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,6144,0.020120888948440552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,5120,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,4096,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,5120,0.01852888862291972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,3072,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,3584,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,3584,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,3584,0.0151235560576121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,2560,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,3072,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,3072,0.007378666765160031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,10240,0.011932444241311816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,2048,0.004600000050332812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,2560,0.007149332927332983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,2560,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,2560,0.013474666410022311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,1536,0.0035617777870761026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,2048,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,2048,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,2048,0.012482666307025485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,1024,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,1536,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,1536,0.005986666513813867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,3584,0.005024888863166173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,1536,0.011058666639857821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,768,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,1024,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,4096,0.007335111498832703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,512,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,1024,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,4096,0.016192000773217943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,768,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,768,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,256,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,768,0.009736888938479954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,512,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,128,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,512,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,256,0.0052933332820733385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,3072,0.014141332772043018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,256,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,64,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,256,32,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,128,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,128,0.00489688871635331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,65536,0.011895111037625207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,128,0.007672000262472365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,64,0.004586666822433472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,32,0.004587555511130227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,51200,0.010057777994208867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,65536,0.0431182218922509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,65536,0.1291511058807373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,51200,0.034612443712022566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,256,1024,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,16384,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,51200,0.037343111303117536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,51200,0.10208355718188816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,12288,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,16384,0.015674667225943673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,16384,0.014767110347747803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,16384,0.038753777742385864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,10240,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,12288,0.013280889226330651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,256,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,12288,0.031194667021433514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,8192,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,10240,0.012089778151777057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,10240,0.010813333094120026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,10240,0.027443556321991816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,65536,0.04620088802443611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,7168,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,8192,0.010456889039940305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,8192,0.009579555855857002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,6144,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,8192,0.02369511127471924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,7168,0.009735110733244155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,7168,0.009335111412737105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,5120,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,7168,0.02198311189810435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,6144,0.009094222552246517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,6144,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,4096,0.004552000098758274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,6144,0.02027555637889438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,5120,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,5120,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,3584,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,512,0.009207111265924241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,4096,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,12288,0.012063110868136087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,4096,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,3072,0.00452977791428566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,3584,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,3584,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,2560,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,3584,0.01512888901763492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,3072,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,3072,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,2048,0.004300444490379757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,3072,0.014373333917723762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,2560,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,2560,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,1536,0.003556444413132138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,2560,0.013365333278973898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,2048,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,2048,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,1024,0.002834666727317704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,2048,0.01238666640387641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,1536,0.0063288890653186375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,1536,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,768,0.0026062221990691293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,1024,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,1536,0.011029333704047732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,1024,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,512,0.002493333278430833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,1024,0.01033688916100396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,768,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,256,0.002001777705219057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,5120,0.018457776970333524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,768,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,768,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,512,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,128,0.0018684443914228016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,512,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,4096,0.016504888733228047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,256,0.005623111294375525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,512,0.009296889106432596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,256,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,128,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,65536,0.008983111215962304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,256,0.009134222235944536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,51200,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,128,0.004534222185611725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,16384,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,12288,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,128,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,64,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,10240,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,128,32,0.005076444397370021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,8192,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,7168,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,65536,0.04298399885495504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,51200,0.035135110219319664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,6144,0.004839111119508743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,5120,0.004566222015354368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,12288,0.012053333222866058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,4096,0.0041982221106688184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,10240,0.011039111349317761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,8192,0.009857777920034196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,3584,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,7168,0.009034666750166152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,3072,0.004237333519591225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,5120,0.007768889268239339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,2560,0.004543999830881755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,4096,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,2048,0.0041893331540955436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,3584,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,1536,0.003536000019974179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,3072,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,1024,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,2560,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,2048,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,768,0.002518222149875429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,512,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,128,32,0.0018133332745896445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,256,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,1024,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,128,0.0018746666610240936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,768,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,64,0.0017546665751271779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,256,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,64,32,0.001806222316291597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,16384,0.01443911095460256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,65536,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,64,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,51200,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,32,0.004543999830881755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,6144,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,16384,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,12288,0.004589333302444882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,65536,0.04310133390956455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,10240,0.004856889032655292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,51200,0.03462488783730401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,8192,0.004184000194072723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,16384,0.014782221780882942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,12288,0.012034667034943899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,10240,0.010793777803579966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,6144,0.00453599997692638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,1536,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,8192,0.009650666680600908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,5120,0.004652444273233414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,4096,0.004238222208287981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,7168,0.009033777647548253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,3584,0.004535111288229625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,512,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,6144,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,3072,0.004237333519591225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,5120,0.00793599999613232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,4096,0.007355555891990662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,64,128,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,3584,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,1536,0.0032711111836963226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,3072,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,1024,0.0027991111079851785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,2560,0.006302222195598815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,768,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,2048,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,512,0.0023457778410779103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,1536,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,256,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,128,0.0018737777653667664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,1024,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,64,0.0015991110768583084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,768,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,32,0.0017928888814316855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,7168,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,512,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,256,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,128,0.004564444637960858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,64,0.004584889031118817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,32,32,32,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,2560,0.0047404443224271136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,16384,0.3654204474555121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,32,32,2048,0.004547555413511065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,16384,0.1635226673550076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,51200,0.4890729056464301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,12288,0.27286577224731445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,16384,0.15625511275397405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,10240,0.22842044300503203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,12288,0.1224035554462009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,12288,0.11576710806952582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,8192,0.18488177988264295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,10240,0.1032480001449585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,10240,0.10741600063112046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,8192,0.07081600030263265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,8192,0.08858578072653876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,7168,0.07855555746290419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,7168,0.0902435580889384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,6144,0.13864089383019343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,51200,0.4612879753112793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,5120,0.11490844355689155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,6144,0.0798106657134162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,6144,0.05524888965818617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,4096,0.0942080020904541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,5120,0.07208622164196439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,4096,0.03923022084765964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,4096,0.058895111083984375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,3072,0.07172622283299764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,3584,0.03524622321128845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,3584,0.054126222928365074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,51200,1.115608851114909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,2560,0.05961955255932278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,3072,0.044406221972571484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,3072,0.03137422270245022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,2048,0.047213332520590894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,2560,0.027246221899986267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,7168,0.15896177291870117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,1536,0.03669866588380601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,2048,0.03354933195643955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,1536,0.018951111369662814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,2560,0.038570665650897555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,1536,0.026030222574869793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,768,0.01942311061753167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,2048,0.02332266668478648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,1024,0.015311110350820752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,1024,0.018545778261290658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,768,0.013179555535316467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,768,0.01585244470172458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,256,0.008030222521887885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,512,0.011824888487656912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,512,0.012200000385443369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,128,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,256,0.0103502223889033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,256,0.009338666995366415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,64,0.0035386664999855887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,128,0.010618666807810465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,128,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,32,0.003798222376240624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,64,0.010384889112578498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,1024,0.02682755556371477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,65536,5120,0.05857866340213352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,512,0.013386666774749756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,65536,0.5189910994635688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,65536,1.0986853705512154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,51200,0.39228622118632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,51200,0.8536399735344781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,16384,0.2780017852783203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,65536,32,0.009777777724795872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,16384,0.13104621569315592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,12288,0.20812977684868705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,16384,0.1244586706161499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,65536,3584,0.08215377728144328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,10240,0.17600977420806885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,65536,0.43607110447353786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,12288,0.08377599716186523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,8192,0.14046666357252333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,51200,0.3616195519765218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,10240,0.08414933416578506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,7168,0.12313600381215413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,12288,0.09843644168641831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,8192,0.05934933159086439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,8192,0.07069866524802314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,6144,0.1055377788013882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,7168,0.0722035566965739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,7168,0.05177866750293308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,6144,0.04510666595564949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,4096,0.0730337765481737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,6144,0.06631466415193346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,3584,0.06355644596947564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,5120,0.060235553317599826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,3072,0.05570666657553779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,3584,0.04351644383536445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,3584,0.029491557015313044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,4096,0.051445331838395864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,2560,0.04665866825315687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,3072,0.03784622086419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,3072,0.025880888104438782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,2048,0.037581334511439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,2560,0.023006222314304773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,2560,0.03354933195643955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,1536,0.029436445898479883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,2048,0.019493333167499967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,2048,0.027408000495698717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,1024,0.020976000361972388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,10240,0.08363999923070271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,5120,0.08965244558122422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,1536,0.02162488963868883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,768,0.015454222758611044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,1024,0.013373333546850415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,1024,0.015452444553375244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,768,0.012364444633324942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,5120,0.03852533300717672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,256,0.006867555694447623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,512,0.0103511114915212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,128,0.0038444445365005066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,768,0.013295999831623502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,256,0.009398221969604492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,512,0.010703110860453712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,64,0.0031840000301599503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,32,0.0033928888539473214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,128,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,128,0.007050666544172499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,64,0.009015999734401703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,32,0.008902221918106079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,65536,0.331657780541314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,65536,0.1777448919084337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,1536,0.01644977761639489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,51200,512,0.011041777829329172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,16384,0.08426400025685628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,65536,0.1561839977900187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,51200,0.13125422265794542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,12288,0.06318933433956571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,51200,4096,0.03285244438383315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,16384,0.04826755656136406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,51200,0.13380266560448542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,16384,0.048398223188188344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,10240,0.053845332728491895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,51200,256,0.007674666742483775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,8192,0.044093334012561373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,12288,0.03353688783115811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,12288,0.036554667684766985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,7168,0.03745955559942458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,10240,0.03626755542225308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,10240,0.031735112269719444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,6144,0.033048888047536217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,8192,0.02459999918937683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,5120,0.028463999430338543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,7168,0.02787999974356757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,7168,0.029430223835839167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,6144,0.019875556230545044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,4096,0.024119999673631456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,6144,0.026995556222067937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,5120,0.017901332841979135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,51200,0.256984896130032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,5120,0.023810666468408372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,4096,0.015565334094895257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,3072,0.018206222189797294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,4096,0.0203493336836497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,2560,0.015448888142903646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,3072,0.01308444473478529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,3584,0.014544000228246054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,3584,0.01871200071440803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,3072,0.01632355484697554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,2048,0.0124177775449223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,1536,0.00998399986161126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,2560,0.012082666986518435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,2048,0.010966222319338055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,1024,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,2560,0.014648000399271647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,2048,0.013120888835854001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,1536,0.0092622223827574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,768,0.006191111273235745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,1536,0.011344888971911537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,512,0.004330666528807746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,1024,0.007889777421951294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,768,0.007540444533030192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,768,0.007510222494602203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,256,0.0028080000645584534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,512,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,128,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,512,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,256,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,64,0.0022026666750510535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,16384,128,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,256,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,32,0.0022231110682090125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,128,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,64,0.006037333359320958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,32,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,65536,0.2530284457736545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,65536,0.14937867058648005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,51200,0.20059733920627168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,16384,0.07063733206854926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,65536,0.13415288925170898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,12288,0.055361777544021606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,51200,0.10717688666449653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,16384,0.037350220812691584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,16384,0.04152355591456095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,10240,0.045781334241231285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,12288,0.029789331886503432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,12288,0.03201333350605435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,16384,3584,0.02029600077205234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,8192,0.03820800118976169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,10240,0.02790133489502801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,7168,0.033709334002600774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,8192,0.02197866638501485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,8192,0.024549333585633173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,6144,0.029607110553317603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,7168,0.020049777295854356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,5120,0.02556177808178796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,7168,0.026330666409598455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,6144,0.01819022165404426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,4096,0.019676445258988273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,6144,0.023836443821589153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,5120,0.016174222032229107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,3584,0.016851555969980028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,5120,0.020850666695170935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,51200,0.11309688621097141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,4096,0.014135999812020196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,3072,0.01482399966981676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,4096,0.01795288920402527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,3584,0.013367999758985309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,8192,0.027572444743580286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,2560,0.01242844429281023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,2048,0.010674667027261523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,3072,0.01216533366176817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,3584,0.01695999999841054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,1536,0.008565333154466417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,2560,0.011296889020336999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,2048,0.009875555833180746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,2560,0.013536888692114087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,10240,0.03142933381928338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,2048,0.012323555847009024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,1024,0.0069004446268081665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,768,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,1536,0.00852888905339771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,1536,0.010326222413116032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,1024,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,768,0.006992888947327931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,768,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,256,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,512,0.006711999989218182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,512,0.006461333483457565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,128,0.002528888897763358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,256,0.006411555740568373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,256,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,64,0.0021991110924217436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,128,0.0052871112194326185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,12288,128,0.0064693333374129395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,16384,1024,0.00886399961180157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,32,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,64,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,65536,0.21828088495466444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,32,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,65536,0.14174044132232666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,51200,0.17361156145731607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,12288,512,0.003953777667548922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,65536,0.12636178069644505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,1024,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,51200,0.1075964437590705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,16384,0.06031199958589342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,12288,0.04901599884033203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,10240,0.039713776773876615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,51200,0.10526044501198663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,16384,0.03865422142876519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,12288,0.027580445011456806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,8192,0.03450399968359206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,12288,0.03072533342573378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,10240,0.027929776244693335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,7168,0.029502222935358684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,10240,0.02666666607062022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,8192,0.02089777754412757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,8192,0.022978666755888198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,7168,0.021724444296624925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,7168,0.023198222120602924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,12288,3072,0.0155404441886478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,6144,0.017309332887331646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,6144,0.019924443629052904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,4096,0.017655111021465726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,3584,0.015059555570284525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,5120,0.015577778220176697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,4096,0.01385244478782018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,5120,0.017916444275114272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,3072,0.013399110900031196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,4096,0.015792000624868605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,2560,0.011448000040319232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,3584,0.012870222330093384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,3584,0.014450665977266101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,2048,0.009742221898502773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,3072,0.011843555503421359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,2560,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,1536,0.008197333249780867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,2048,0.009123555488056606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,2560,0.012164444559150271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,16384,0.03979644510481093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,1024,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,2048,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,768,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,1536,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,1024,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,1536,0.009062222308582729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,512,0.0038986665507157645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,768,0.0069288888739215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,6144,0.025616000096003216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,1024,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,512,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,128,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,768,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,5120,0.02202133337656657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,512,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,256,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,64,0.0021564445147911706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,32,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,256,0.005480888817045424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,10240,128,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,128,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,64,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,32,0.005689777847793367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,51200,0.13809777630700007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,65536,0.09118400017420451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,65536,0.07019378079308404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,16384,0.04992977778116862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,10240,3072,0.013231111069520315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,12288,0.03609155615170797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,51200,0.06971466541290283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,51200,0.059227552678849965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,16384,0.022759111391173467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,10240,0.026965333355797663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,16384,0.027647111150953505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,8192,0.022656889425383672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,12288,0.01868888901339637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,10240,0.019018666611777413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,10240,256,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,8192,0.014709333578745524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,7168,0.019860444797409903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,6144,0.017533333765135873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,7168,0.01516266663869222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,7168,0.01756355497572157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,5120,0.015096889601813423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,6144,0.012730666332774691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,6144,0.015792888071801927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,65536,0.17416621579064262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,4096,0.013260444833172692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,5120,0.011689778003427716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,3584,0.01109777804878023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,5120,0.013730666703648038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,4096,0.010660444696744284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,3072,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,4096,0.012871111432711283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,3584,0.009677333136399588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,2560,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,3584,0.011771555576059552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,3072,0.00869955536392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,3072,0.010974222587214576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,10240,0.017871111631393433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,2048,0.007558222446176741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,2560,0.00829777783817715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,1536,0.006224000205596288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,2560,0.010300444232092964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,2048,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,1024,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,8192,0.016135111451148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,1536,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,2048,0.008962666822804345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,1536,0.007272889216740926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,512,0.0028008888992998335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,1024,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,768,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,256,0.0025271111064487034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,768,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,512,0.006024888820118374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,128,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,512,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,256,0.0058462222417195635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,64,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,32,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,128,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,128,0.004955555415815777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,65536,0.15075022644466823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,32,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,12288,0.021653332644038733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,51200,0.11916800340016682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,65536,0.08385688728756374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,16384,0.041963554090923734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,65536,0.06646311283111572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,8192,768,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,51200,0.05362488826115926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,51200,0.0631564458211263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,16384,0.021911111142900255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,10240,0.027816000911924574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,16384,0.025783111651738484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,12288,0.01761244402991401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,8192,0.023680888944202002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,10240,0.01608622239695655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,12288,0.020000000794728596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,7168,0.017848889032999676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,10240,0.017277333471510146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,8192,0.014122666584120857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,6144,0.015840889679061044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,8192,0.015288000305493673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,7168,0.013387555877367655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,64,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,5120,0.01347733371787601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,7168,0.016748444901572335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,4096,0.011567999919255575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,6144,0.012054222325483957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,5120,0.011094222466150919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,6144,0.01608622239695655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,5120,0.013452444639470843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,4096,0.009723555710580613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,3072,0.009309333231714036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,4096,0.012159999873903064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,3584,0.00869688888390859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,3584,0.011515555282433828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,3072,0.007918222082985772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,12288,0.03293955657217238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,2048,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,3072,0.010952889091438718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,2560,0.00758844448460473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,1536,0.0057795556883017225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,2560,0.009727111293209923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,2048,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,1024,0.0041582224269707995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,2048,0.008096888661384583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,8192,256,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,1536,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,768,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,1536,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,1024,0.006367110957702001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,512,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,768,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,768,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,256,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,3584,0.010227555202113258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,512,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,128,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,256,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,256,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,64,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,2560,0.008034666379292807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,128,0.0058266665372583605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,8192,1024,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,65536,0.12542933887905544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,64,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,32,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,51200,0.09903822342554729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,65536,0.06389422549141778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,65536,0.07667733563317193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,1024,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,16384,0.03582133187188043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,7168,512,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,51200,0.051593777206208974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,16384,0.020963556236690942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,12288,0.028526223368114893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,16384,0.023972445064120825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,10240,0.024314666787783306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,12288,0.017252445220947266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,12288,0.017934223016103108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,10240,0.015393777026070489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,7168,32,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,7168,128,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,10240,0.015957333975368075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,7168,0.018461333380805123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,8192,0.01351111133893331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,6144,0.01604977746804555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,8192,0.014456889695591398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,7168,0.012717333104875354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,5120,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,7168,0.015473778049151102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,6144,0.01166311071978675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,4096,0.01162577751610014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,6144,0.014136888914638095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,5120,0.01070666644308302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,3584,0.009480000370078618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,5120,0.013286222186353473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,4096,0.00924711094962226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,3072,0.00850844466023975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,4096,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,2560,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,3072,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,3584,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,51200,0.05704889032575819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,2048,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,3072,0.010041777458455827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,1536,0.005493333356248007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,2560,0.007346666521496243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,2048,0.007256000406212277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,1024,0.0038737778862317405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,2560,0.008775110873911116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,768,0.0032266666077905228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,2048,0.007730666961934831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,1536,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,8192,0.02038755516211192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,512,0.002776888923512565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,1024,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,1536,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,256,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,768,0.006011555592219035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,1024,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,512,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,256,0.005714666512277391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,512,0.005373333477311664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,64,0.001826666709449556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,256,0.004942222187916438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,128,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,128,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,65536,0.10770755343967015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,64,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,32,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,51200,0.08533333407508002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,6144,3584,0.008634666601816813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,65536,0.06776355372534858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,16384,0.031699554787741765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,65536,0.06133244434992472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,51200,0.05171644356515673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,51200,0.049647112687428795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,16384,0.020242666204770405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,10240,0.021743999587164983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,16384,0.02134399943881565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,12288,0.01663733356528812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,8192,0.019307555423842538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,10240,0.014917333920796713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,128,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,10240,0.014951111541854011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,6144,768,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,7168,0.016520889268981088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,6144,0.014858666393491956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,6144,32,0.0019288888821999233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,8192,0.013783111340469785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,7168,0.01240355521440506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,5120,0.01310222182008955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,7168,0.014712888333532544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,6144,0.01129066695769628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,4096,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,6144,0.014533332652515836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,5120,0.009837333526876237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,3584,0.008667555948098501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,4096,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,5120,0.012296888563368054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,3072,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,4096,0.011350221931934357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,3584,0.008199999729792276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,12288,0.02569866677125295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,2560,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,3584,0.010293333066834344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,3072,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,2048,0.006068444500366847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,3072,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,2560,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,1536,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,2560,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,2048,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,1024,0.0041911109454101985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,1536,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,2048,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,768,0.003539555602603488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,1536,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,1024,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,512,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,8192,0.01308444473478529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,768,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,256,0.002463999928699599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,1024,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,128,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,768,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,512,0.005767111149099138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,64,0.0018977777411540349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,256,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,512,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,5120,32,0.0021395554973019496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,256,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,128,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,65536,0.08783644437789917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,64,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,32,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,51200,0.06945866346359253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,65536,0.0611048870616489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,16384,0.02499377727508545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,65536,0.05923555294672648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,12288,0.019688889384269714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,51200,0.04770222306251526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,51200,0.04707022176848518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,16384,0.01959111127588484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,10240,0.017179555363125272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,12288,0.015984000431166757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,8192,0.014904888139830695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,12288,0.015498666299713982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,10240,0.014471999473041959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,7168,0.01312711089849472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,8192,0.012503110700183444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,8192,0.012390221986505719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,6144,0.01146044416560067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,7168,0.011703111231327057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,7168,0.014452444182501899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,5120,128,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,5120,0.010096888575288985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,4096,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,6144,0.010614222122563256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,5120,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,6144,0.01238933371173011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,3584,0.007925333248244392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,5120,0.011976888610257043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,4096,0.008346666892369589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,3072,0.007236444287829929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,4096,0.010943111446168689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,3584,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,2560,0.006320000108745363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,3584,0.009208888643317753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,3072,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,2048,0.005403555515739653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,3072,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,2560,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,5120,12288,0.01651022169325087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,16384,0.01924088928434584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,2560,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,2048,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,1024,0.00319377767542998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,768,0.002834666727317704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,2048,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,1536,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,1024,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,10240,0.013804444836245643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,512,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,1536,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,256,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,768,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,1024,0.006142222219043308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,128,0.0021271111650599376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,768,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,512,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,256,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,64,0.0019617778145604664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,32,0.002053333239422904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,512,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,256,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,4096,128,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,128,0.004946666873163647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,65536,0.0759084423383077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,51200,0.059966219796074756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,32,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,65536,0.05828444163004557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,16384,0.02308977809217241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,65536,0.05831466780768501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,51200,0.046880000167422824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,12288,0.01856355534659492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,16384,0.01923733287387424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,16384,0.0184560004207823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,10240,0.015855110353893705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,12288,0.015841777125994366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,4096,1536,0.0041680000722408295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,8192,0.013795555465751223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,12288,0.014884443746672736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,10240,0.014167111780908374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,7168,0.012896888785892062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,8192,0.012424888710180918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,10240,0.013442666994200813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,6144,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,8192,0.012432888978057437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,7168,0.011188444164064197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,5120,0.009642666412724389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,7168,0.012784889174832238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,6144,0.009750222166379293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,4096,0.009003555609120263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,5120,0.008981333010726506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,3584,0.007601777712504069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,5120,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,4096,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,3072,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,4096,64,0.005192000004980299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,3584,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,2560,0.006056889063782162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,3584,0.008646222452322641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,3072,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,2048,0.005011555635266834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,51200,0.04537422127193875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,3072,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,2560,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,2560,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,2048,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,1024,0.003157333367400699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,2048,0.006923555499977536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,1536,0.006735999965005451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,1536,0.0063546668324205614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,768,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,1024,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,512,0.0025351111673646504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,768,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,1024,0.005680888891220093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,256,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,768,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,512,0.005717333406209946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,6144,0.013436444103717804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,128,0.0021955555097924336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,512,0.005295111073387994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,256,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,64,0.0018737777653667664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,32,0.0019617778145604664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,256,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3584,128,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,4096,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,65536,0.06618933545218574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,128,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,51200,0.05097422334882948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,16384,0.019703111714786954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,32,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,65536,0.0571920010778639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,65536,0.05562577644983927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3584,1536,0.003927999900446998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,51200,0.045985778172810875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,51200,0.043567111094792686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,16384,0.018971555762820773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,12288,0.015475556254386902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,8192,0.012228444218635559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,10240,0.013824000126785703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,12288,0.014506666196717156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,7168,0.011375111010339526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,8192,0.011543999943468304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,8192,0.011786667009194693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,6144,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,7168,0.01073688848151101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,7168,0.012101333174440594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,5120,0.009580444130632613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,6144,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,6144,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,4096,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,5120,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3584,64,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,5120,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,3584,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,4096,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,12288,0.016765332884258695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,3072,0.00701422244310379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,10240,0.014167111780908374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,3584,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,2560,0.007368889119890001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,16384,0.01780177818404304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,3584,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,3072,0.007660444411966536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,2048,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,3072,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,1536,0.0047600000268883174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,2048,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,2560,0.006674666785531574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,1024,0.0035457776652442086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,2048,0.0063582224150498705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,1536,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,768,0.003162666741344664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,1024,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,1024,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,512,0.0028320000403457214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,768,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,768,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,256,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,512,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,512,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,4096,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,256,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,256,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,64,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,128,0.00563733321097162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,32,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,128,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,3072,2560,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,65536,0.05907289187113444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,64,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,51200,0.0455075568623013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,32,0.005316444569163852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,1536,0.005998222364319696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,65536,0.05365955498483446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,16384,0.01993955506218804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,51200,0.04519288738568624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,51200,0.04215911030769348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,16384,0.018569777409235638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,10240,0.014166222678290473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,16384,0.017229333519935608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,12288,0.014817777607176038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,8192,0.013184888495339287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,12288,0.014152889450391134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,10240,0.01277599980433782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,7168,0.011634666886594562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,10240,0.012730666332774691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,8192,0.011698666546079846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,8192,0.0114720000161065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,7168,0.010236444572607676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,3072,10240,0.01309511148267322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,7168,0.011152889165613385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,5120,0.008824889030721452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,6144,0.009308444129096137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,6144,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,4096,0.007968000239796108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,5120,0.008957333034939235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,65536,0.05624000231424967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,5120,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,4096,0.008031999899281396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,3072,0.006678222368160884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,4096,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,3584,0.007754666937722101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,12288,0.01627377834584978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,2560,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,3584,0.00720355576939053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,2048,0.005015999906592899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,3072,0.0069146665434042616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,2560,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,1536,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,2560,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,3072,128,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,2048,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,1024,0.003572444534964032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,2048,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,768,0.0031840000301599503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,6144,0.010434666441546546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,1536,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,512,0.002847111059559716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,1024,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,768,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,768,0.0053048887186580235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,256,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,512,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,128,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,512,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,256,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,3584,0.007185777856243982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,64,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,256,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,128,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,65536,0.047071112526787647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,64,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,3072,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,32,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,51200,0.03679644399219089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,65536,0.0555573304494222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,1536,0.006665777828958299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,51200,0.04045333464940389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,51200,0.04474577638838026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2560,1024,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,12288,0.013635555903116861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,16384,0.018292443619834054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,16384,0.0166311115026474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,10240,0.012159110771285163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,12288,0.014017777310477363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2560,32,0.0018133332745896445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,10240,0.012079111403889127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,8192,0.011590222517649332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2560,128,0.004861333303981357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,10240,0.012420444852775998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,7168,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,8192,0.01090666651725769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,6144,0.00941244430012173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,8192,0.010808000134097205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,7168,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,7168,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,5120,0.008622222476535374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,6144,0.009422221945391761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,4096,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,6144,0.009030222064918941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,5120,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,16384,0.014631999863518609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,3584,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,5120,0.008303999900817871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,3072,0.005758222192525864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,4096,0.008079999850855933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,4096,0.0075084442893664045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,3072,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,12288,0.013909333282046847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,3072,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,2048,0.0042444442709287005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,2560,0.0069857777820693115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,2560,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,1536,0.003846222327815162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,2048,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,1024,0.003144888828198115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,1536,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,2048,0.006032889087994893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,768,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,1536,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,1024,0.006021333237489064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,512,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,1024,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,768,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,65536,0.05134133497873942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,256,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,768,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,128,0.0019128889673286015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,512,0.0058808889653947614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,256,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,512,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,128,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,3584,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,65536,0.03605511122279697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,3584,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,64,0.005098666581842634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,51200,0.02844711144765218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,32,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,16384,0.012607110871209038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,65536,0.05475644601715935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,65536,0.048935112026002675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,51200,0.0444764428668552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,12288,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,51200,0.03901777664820353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,16384,0.01719555589887831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,10240,0.009019555317031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,16384,0.016295111841625638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,8192,0.013256000147925483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,12288,0.013122667041089801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,12288,0.013554666605260638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,10240,0.012059555285506778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,7168,0.009684444301658207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,8192,0.010562666588359408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,6144,0.009233777721722921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,8192,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,7168,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,5120,0.008301333420806462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,7168,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,6144,0.009485333330101436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,2048,256,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,4096,0.007487110793590546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,6144,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,5120,0.008640888664457533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,3584,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,5120,0.008319999608728621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,4096,0.008270222279760573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,3072,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,3584,0.007670222057236566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,2560,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,3072,0.007804444266690149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,3072,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,2560,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,2048,0.004330666528807746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,2560,0.006335999816656113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,2048,0.006713777780532837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,1536,0.0038524443904558816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,10240,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,2048,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,1024,0.003192000091075897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,1536,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,1536,0.005807999935415056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,2048,128,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,1024,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,1024,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,2048,2560,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,768,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,4096,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,256,0.0021964444054497615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,512,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,3584,0.007096000015735626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,512,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,128,0.0020968889196713767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,256,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,64,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1536,128,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,256,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,32,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,128,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,65536,0.02830044428507487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,64,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,51200,0.022094221578703985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,65536,0.05437066819932726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,65536,0.04719644453790453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,16384,0.010600888894663917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,51200,0.04383555716938443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,768,0.0028080000645584534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,51200,0.03783644570244683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,12288,0.008807111117574904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1536,512,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,16384,0.015804444750150044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,16384,0.015125332607163323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,12288,0.013086222112178802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,768,0.005514666852023866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,8192,0.0075911109646161394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,12288,0.012442666623327466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,10240,0.011765333513418833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,7168,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,6144,0.0063031112982167145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,10240,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,7168,0.009932444327407414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,8192,0.010016889207892949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,5120,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,6144,0.009375111096435124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,4096,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,6144,0.00867911097076204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,5120,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,3584,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,5120,0.008269333177142674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1536,32,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,4096,0.007927999728255803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,3072,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,4096,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,2560,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,3584,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,3072,0.007290666302045186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,2048,0.005183111048407025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,3072,0.007167999943097432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,2560,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,1536,0.005137777576843898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,10240,0.008183110919263627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,2048,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,1024,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,2048,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,1536,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,768,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,1536,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,8192,0.010466666685210334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,512,0.002502222235004107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,1024,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,768,0.005640888793600931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,256,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,7168,0.009298666483826106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,768,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,128,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,512,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,512,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,256,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,64,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,1024,32,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,256,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,65536,0.022293332550260756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,128,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,3584,0.007927999728255803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,51200,0.01788444485929277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,64,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,32,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,65536,0.05406755540106031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,16384,0.00958933350112703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,51200,0.043212443590164185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,1024,2560,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,12288,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,51200,0.03757777810096741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,16384,0.015809777710172866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,10240,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,16384,0.014831999937693277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,12288,0.013389333254761167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,12288,0.012158222496509552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,8192,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,10240,0.012073777616024017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,7168,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,10240,0.011039111349317761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,8192,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,6144,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,8192,0.009896000226338705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,7168,0.010008888940016428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,5120,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,7168,0.009331555830107795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,128,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,6144,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,4096,0.00656088896923595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,6144,0.008880889250172509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,5120,0.009000889129108852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,3584,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,4096,0.00794400026400884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,3072,0.005389333185222413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,3584,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,4096,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,2560,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,65536,0.04632088873121473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,3584,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,3072,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,2048,0.004229333251714706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,3072,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,1536,0.0038177776667806837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,2560,0.00693866651919153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,2048,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,1024,0.0031840000301599503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,1024,1024,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,2048,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,768,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,1536,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,1024,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,512,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,1024,0.005633777628342311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,256,0.0022159999029503926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,768,0.005728888842794631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,768,0.005392888767851724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,512,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,256,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,512,0.004956444518433677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,64,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,256,0.005072000126043956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,32,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,128,0.0053093334039052325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,5120,0.008092444803979661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,65536,0.0158924443854226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,64,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,65536,0.04788088798522949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,32,0.004961777892377642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,51200,0.013493333425786761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,65536,0.045772443215052284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,2560,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,51200,0.03847377830081516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,12288,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,51200,0.03683022326893277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,16384,0.014428445034556918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,16384,0.015806222955385845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,10240,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,12288,0.012787555654843649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,8192,0.005693333430422678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,12288,0.012401777837011548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,10240,0.011752888560295105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,7168,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,10240,0.011045333411958484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,768,128,0.0019199999256266488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,8192,0.010703110860453712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,6144,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,8192,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,7168,0.010000000397364298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,5120,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,6144,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,7168,0.009333333207501305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,4096,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,768,128,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,6144,0.008837333156002892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,3584,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,5120,0.0086666668454806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,3072,0.00553955551650789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,4096,0.008194666769769456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,3584,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,4096,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,2560,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,3072,0.007589333587222629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,3584,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,2048,0.004176889028814104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,3072,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,2560,0.007100444700982835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,1536,0.0035626664757728577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,768,1536,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,2048,0.006918222126033571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,1536,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,2048,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,1536,0.005633777628342311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,768,0.002833777831660377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,1024,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,1024,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,512,0.0025128889828920364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,768,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,768,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,256,0.0021537778278191886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,512,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,512,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,128,0.001828444500764211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,256,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,5120,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,64,0.0018088887963030075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,256,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,512,128,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,128,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,65536,0.017083555459976196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,64,0.004951111144489712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,51200,0.014476444986131458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,32,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,65536,0.046624888976415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,16384,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,65536,0.04421866602367825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,51200,0.037295109695858426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,12288,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,16384,0.015823110938072205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,16384,0.014488889111412896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,10240,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,12288,0.013276444541083442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,12288,0.012063110868136087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,8192,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,10240,0.012010667059156628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,7168,0.0053173332578606075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,10240,0.011128889189826118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,8192,0.010426667001512317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,8192,0.009661333428488838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,6144,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,16384,0.011042666931947073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,7168,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,5120,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,7168,0.009285333255926767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,4096,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,6144,0.009328888522254096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,5120,0.008658666577604082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,6144,0.008286221987671321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,512,2560,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,5120,0.008522666990756989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,4096,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,3072,0.005429333282841577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,4096,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,3584,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,2560,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,3072,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,3584,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,2048,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,51200,0.034978667894999184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,3072,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,1536,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,2560,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,1024,0.00300177786913183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,2048,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,2560,0.006426666759782367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,1536,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,2048,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,1536,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,512,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,1024,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,768,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,256,0.0021217777911159727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,768,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,512,0.005884444548024072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,128,0.001828444500764211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,512,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,256,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,64,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,256,0.0047528888616296975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,32,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,3584,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,512,1024,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,128,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,51200,0.010325333310498131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,64,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,65536,0.04586044285032484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,256,32,0.004581333448489507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,16384,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,65536,0.04211466511090597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,51200,0.03707911239729987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,12288,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,51200,0.034501334031422935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,16384,0.015491555134455362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,256,768,0.0026364444444576898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,256,1024,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,12288,0.01239733315176434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,8192,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,10240,0.011694221860832639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,10240,0.010838222172525195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,7168,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,8192,0.01071111112833023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,8192,0.009941333697901832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,6144,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,7168,0.010199999643696679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,7168,0.009308444129096137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,5120,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,65536,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,6144,0.008519111408127679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,4096,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,5120,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,5120,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,3584,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,4096,0.008007999923494127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,10240,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,4096,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,3072,0.004592888885074192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,16384,0.014156444205178155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,3584,0.006917333437336816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,2560,0.00453955555955569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,3072,0.007601777712504069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,2048,0.004522666749027041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,2560,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,3072,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,1536,0.003532444437344869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,2560,0.0064071110553211635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,2048,0.006562666760550604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,1024,0.002879111096262932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,2048,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,1536,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,768,0.0026542221506436667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,1536,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,6144,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,512,0.002456888970401552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,1024,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,768,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,256,0.0021359999146726397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,768,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,512,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,128,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,256,0.00527555536892679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,64,0.0017946666727463405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,512,0.0053022222386466126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,128,32,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,256,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,65536,0.009045333498054082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,128,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,3584,0.00758844448460473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,51200,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,64,0.004592888885074192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,16384,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,12288,0.005199111170238919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,65536,0.042428443829218544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,10240,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,51200,0.03392444385422601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,16384,0.0141004439857271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,8192,0.005197333378924264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,12288,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,7168,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,6144,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,10240,0.011071999867757162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,5120,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,4096,0.004230222354332606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,8192,0.009575111170609793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,7168,0.009093333449628618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,3584,0.0046968890560997855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,6144,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,5120,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,2560,0.004943111290534337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,3584,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,128,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,2048,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,3072,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,1536,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,1024,0.0028719999310043124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,17,128,12288,0.013099555340078143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,2560,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,768,0.0025048889219760895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,512,0.0021582223061058256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,128,32,0.004947555561860402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,2048,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,256,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,128,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,1024,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,768,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,32,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,512,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,256,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,65536,0.00865955568022198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,128,0.004959110998445087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,64,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,32,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,65536,0.04192089041074117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,64,3072,0.004514666481150521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,12288,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,51200,0.034274667501449585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,10240,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,8192,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,4096,0.006979555719428592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,16384,0.014675555957688225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,7168,0.00519822206762102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,12288,0.011931555966536203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,6144,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,8192,0.009517333573765224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,5120,0.004570666700601578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,7168,0.009325332939624786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,4096,0.00423911131090588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,64,1536,0.005624889085690181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,6144,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,3584,0.004193777839342753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,3072,0.0041857775714662345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,5120,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,2560,0.004281777888536453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,4096,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,2048,0.004535111288229625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,3584,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,3072,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,16384,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,1024,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,2560,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,768,0.0024622221373849446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,2048,0.006246222390068903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,512,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,1536,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,256,0.0021226666867733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,1024,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,128,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,768,0.005026666654480828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,64,0.00152800004515383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,10240,0.0110880004035102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,32,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,128,0.004863111095296012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,64,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,32,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,51200,0.007669332954618666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,17,32,1536,0.003506666670242945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,51200,0.4898008770412869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,16384,0.3235849009619819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,51200,0.9777395460340711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,512,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,17,32,256,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,16384,0.1636319955190023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,12288,0.2262942261166043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,12288,0.12189688947465684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,16384,0.13858311706119114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,10240,0.1881937715742323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,12288,0.12363910675048828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,10240,0.1035733355416192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,8192,0.15430222617255315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,51200,0.45942311816745335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,10240,0.0948808921708001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,8192,0.08805155754089355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,7168,0.13258222738901773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,12288,0.20578932762145996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,10240,0.1706684430440267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,7168,0.09058666891521877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,16384,0.2742488914065891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,6144,0.11485421657562256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,7168,0.06906577613618639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,8192,0.14052266544765896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,5120,0.09639644622802734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,6144,0.07739377684063382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,6144,0.05566133393181694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,4096,0.08037422100702922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,7168,0.14830488628811306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,5120,0.04698310958014595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,8192,0.0710044436984592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,6144,0.1575733290778266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,4096,0.05918400155173408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,4096,0.03945688737763299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,5120,0.11899999777475993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,3072,0.05948444207509359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,4096,0.08983644511964585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,3584,0.03527644607755873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,3584,0.05238755544026693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,2560,0.05018488897217644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,3584,0.08354577753278945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,3072,0.031382222970326744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,3072,0.06708533234066434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,2560,0.03952799903021918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,2048,0.04085510969161987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,2560,0.0273671117093828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,5120,0.06944177548090617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,1536,0.03158311049143473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,2560,0.058168000645107694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,2048,0.023203555080625746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,2048,0.03388266762097677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,1024,0.022639999787012737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,3584,0.06843822532229953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,1536,0.01891288823551602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,1536,0.04161422120200263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,1024,0.018838221828142803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,2048,0.05039111110899183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,1024,0.0151591118839052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,1024,0.03266133202446832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,3072,0.05099822084108988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,768,0.013245333400037555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,768,0.01640977793269687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,768,0.028135110934575398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,512,0.012496888637542725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,256,0.007196444604131911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,512,0.012657778130637275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,512,0.02446222139729394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,51200,0.8219600253634982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,128,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,256,0.01053600013256073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,256,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,256,0.020273778173658583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,64,0.004462222258249919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,128,0.010497777826256223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,32,0.0046657779150538975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,128,0.015819556183285184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,768,0.017828444639841717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,64,0.01035733355416192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,1536,0.025153777665562097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,65536,512,0.0125928885406918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,65536,32,0.010399111443095738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,51200,0.39129066467285156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,65536,0.514952023824056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,128,0.010334221853150262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,51200,0.8515360090467664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,65536,1.0930577384101021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,16384,0.2769893275366889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,65536,0.4354151089986165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,16384,0.13128356138865152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,12288,0.20750933223300508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,16384,0.11135289404127334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,51200,0.3303031126658122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,12288,0.09828977452384101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,10240,0.17245599958631727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,12288,0.08468711376190186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,10240,0.08347377512190078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,8192,0.1402746703889635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,10240,0.07550399833255343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,8192,0.07099911239412096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,8192,0.058832890457577176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,12288,0.16763999727037218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,16384,0.22518576516045463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,7168,0.1226773394478692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,10240,0.14276977380116782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,6144,0.10473955339855617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,7168,0.07779022057851155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,7168,0.0524871117538876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,5120,0.0888764460881551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,6144,0.04570933183034261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,6144,0.06688177585601807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,7168,0.10882933272255792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,5120,0.058111998769972056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,4096,0.07248266537984212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,5120,0.03870577613512675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,6144,0.1309537755118476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,3584,0.06293066342671712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,4096,0.04840622345606486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,4096,0.03290844294759963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,4096,0.07004533211390178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,3584,0.043062223328484424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,3584,0.029306666718588933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,3072,0.05496977766354879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,3584,0.0686826639705234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,8192,0.11684444215562607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,2560,0.04632622334692213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,3072,0.026365333133273657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,3072,0.055254220962524414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,3072,0.03724977705213759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,2560,0.022993778189023335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,2560,0.032272001107533775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,2560,0.04851288927925957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,2048,0.027348443865776062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,1536,0.02907199992073907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,2048,0.023150222169028387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,5120,0.09713244438171387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,1024,0.020578665865792167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,1536,0.016855110724767048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,51200,0.6667448679606119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,2048,0.042025778028700084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,1536,0.021379555265108745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,768,0.015108444624476962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,1536,0.035700443718168474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,1024,0.016131555040677387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,512,0.011024889018800525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,768,0.013612444202105204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,1024,0.030060443613264296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,768,0.01182755579551061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,256,0.006412444429265128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,65536,0.9031360414293078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,512,0.011729777687125735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,512,0.021175111333529156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,2048,0.03686577743954129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,256,0.010384000009960597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,256,0.00868266655339135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,256,0.017994667092959087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,128,0.009482666850090027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,64,0.0028239999794297745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,128,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,128,0.014154665999942355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,32,0.00282222218811512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,64,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,32,0.009005332986513773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,51200,1024,0.015523556205961438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,65536,0.32293332947625053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,65536,0.172252443101671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,51200,768,0.024544888072543677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,51200,512,0.010738666686746808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,51200,0.2508488761054145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,51200,128,0.0037173334923055437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,51200,0.13137866391075984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,65536,0.16124978330400255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,16384,0.08964533276028103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,51200,0.12381954987843831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,16384,0.04859911070929634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,12288,0.06989066468344794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,16384,0.04327022367053562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,10240,0.05318222112125821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,12288,0.03629511263635423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,12288,0.03665333324008518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,16384,0.05611999829610189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,12288,0.04322488771544563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,8192,0.04361777835422092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,10240,0.0317813323603736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,10240,0.03126310971048143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,7168,0.03810933232307434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,10240,0.03828533490498861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,8192,0.024731554918819006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,8192,0.02753066685464647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,6144,0.033180445432662964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,65536,0.1800675524605645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,8192,0.03423110975159539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,7168,0.02392533255947961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,7168,0.03012622065014309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,51200,0.1384577751159668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,5120,0.028146667612923518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,6144,0.019885333047972787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,6144,0.027056889401541814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,5120,0.01773155563407474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,4096,0.023400889502631292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,5120,0.02414488957987891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,6144,0.02910755409134759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,3584,0.020578665865792167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,7168,0.03285333183076646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,4096,0.01551644504070282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,4096,0.02012800011369917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,3072,0.01763288842307197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,3584,0.014468444718254937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,4096,0.024319110645188227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,3584,0.018231999542978074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,3584,0.020796444680955675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,3072,0.013481777575280933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,3072,0.016178665889634024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,2048,0.01240533341964086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,3072,0.0192275560564465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,2560,0.01240355521440506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,2560,0.014836443795098199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,1536,0.00998577806684706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,2560,0.01720711092154185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,2048,0.011196444431940714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,1024,0.00739555557568868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,1536,0.009591999981138442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,1536,0.014452444182501899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,1536,0.010807111031479306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,768,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,2048,0.016504888733228047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,1024,0.009289777941173976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,5120,0.025694222913848028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,512,0.0041759999261962045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,1024,0.01311466677321328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,768,0.007695111135641734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,768,0.007501333124107785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,256,0.002510222295920054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,768,0.011396444506115384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,2560,0.015051555302408008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,512,0.006991110742092133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,512,0.00657155571712388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,128,0.0022035555707083806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,512,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,256,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,64,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,256,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,16384,32,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,256,0.009528000321653154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,128,0.006789333290523953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,128,0.007671111159854465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,2048,0.012847111456924014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,64,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,32,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,65536,0.2531084484524197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,65536,0.14855199389987522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,51200,0.19789422882927787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,16384,0.06555110878414579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,1024,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,65536,0.13857244120703802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,51200,0.11282755268944634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,12288,0.051128000020980835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,51200,0.10734133587943183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,16384,0.038093331787321306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,16384,0.04115110966894362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,10240,0.043476445807351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,12288,0.032032890452278986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,12288,0.03457422057787577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,12288,0.03684088918897841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,10240,0.027855998939938013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,8192,0.036553776926464505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,10240,0.027335999740494624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,10240,0.033180445432662964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,65536,0.1448497772216797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,7168,0.03162311183081733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,8192,0.021977777282396953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,8192,0.024269334144062463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,8192,0.02949955397182041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,6144,0.027664000789324444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,7168,0.020852444900406733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,7168,0.027842667367723253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,6144,0.02373955481582218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,16384,0.046893331739637584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,5120,0.024121777878867254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,6144,0.018223111828168232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,6144,0.026199999782774184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,4096,0.020668443706300523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,16384,128,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,5120,0.016228444046444364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,5120,0.022240888741281297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,3584,0.018265777164035372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,5120,0.022368000613318548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,4096,0.019613333874278598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,3584,0.016920889417330425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,3584,0.013407111167907715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,3072,0.016184889607959323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,3584,0.018615111708641052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,7168,0.026344888740115698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,2560,0.011693333586057028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,3072,0.012411555482281579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,3072,0.015311999453438653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,3072,0.01718222267097897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,2048,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,2560,0.013767999907334646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,2560,0.011390222443474663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,2560,0.015726221932305228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,1536,0.008254222571849823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,2048,0.009582222335868413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,2048,0.012087110843923358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,1024,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,51200,0.11356266339619954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,2048,0.015557333827018738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,1536,0.008605333666006723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,1536,0.01013422260681788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,768,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,1024,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,1536,0.013745778136783175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,1024,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,512,0.0038977778620190094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,1024,0.012422222230169507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,768,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,256,0.0028888889484935333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,768,0.010341333018408881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,512,0.0069288888739215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,512,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,128,0.002456888970401552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,512,0.009829333258999718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,256,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,256,0.005628444254398346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,64,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,256,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,128,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,128,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,128,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,64,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,32,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,4096,0.018575999471876357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,51200,0.17283911175198027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,12288,768,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,65536,0.128839996125963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,51200,0.10770133468839858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,16384,0.05860177675882975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,51200,0.10024622413847183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,12288,32,0.002191999927163124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,16384,0.039573334985309176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,65536,0.13247733645968968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,16384,0.039064887497160174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,12288,0.04551288816663954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,65536,0.21784178415934244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,16384,0.04329244295756022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,10240,0.03753688931465149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,51200,0.1066933340496487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,12288,0.028007109959920246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,12288,0.030753778086768255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,65536,0.13941333029005262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,12288,0.03528177738189697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,10240,0.02536533276240031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,10240,0.026518222358491685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,7168,0.028031110763549805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,8192,0.021165332860416834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,10240,0.031587554348839655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,8192,0.022980444961124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,6144,0.02471911079353756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,8192,0.028205331828859117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,7168,0.023189332750108507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,7168,0.019911110401153564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,5120,0.021252445048756067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,7168,0.025898666845427618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,6144,0.01717688971095615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,6144,0.020124443703227572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,4096,0.017814222309324477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,5120,0.018377777602937486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,5120,0.015443555182880826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,6144,0.02325155503220028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,5120,0.02108088963561588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,3584,0.015850666496488783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,4096,0.014616888430383472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,4096,0.013758222262064615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,3072,0.014197332991494073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,4096,0.01953155630164676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,3584,0.01314133322901196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,8192,0.03132177723778619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,3584,0.014244443840450711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,2560,0.011013333168294696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,3584,0.017802667286660936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,3072,0.012010667059156628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,2048,0.009678222239017487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,2560,0.010475555227862464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,3072,0.013067555924256643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,2560,0.01200355589389801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,3072,0.016716443830066256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,1536,0.008042666647169325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,2560,0.015623999966515435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,2048,0.009667555491129557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,1024,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,2048,0.010984000232484607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,2048,0.014935111006100973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,1536,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,1536,0.009024000002278222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,768,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,1536,0.013056000073750814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,1024,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,512,0.003858666867017746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,1024,0.011598221957683563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,768,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,768,0.010382222632567087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,4096,0.015474667151769003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,256,0.0028231110837724474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,512,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,512,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,128,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,512,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,256,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,64,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,128,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,128,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,10240,32,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,128,0.007644444704055786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,64,0.00572533326016532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,65536,0.17326400015089247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,32,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,51200,0.1308399968677097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,1024,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,10240,768,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,65536,0.07230933507283528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,16384,0.044450667169359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,51200,0.06930133369233873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,51200,0.05836711327234904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,12288,0.03528711199760437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,16384,0.022592888938056097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,16384,0.027560888065232172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,16384,0.04240622123082479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,256,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,12288,0.02162488963868883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,12288,0.020261334048377145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,65536,0.1314097775353326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,51200,0.1055484414100647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,8192,0.022967111733224656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,12288,0.03453333179155985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,10240,0.01737333337465922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,10240,0.019137778215938143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,8192,0.014800000521871777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,10240,0.03130311104986403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,8192,0.01718488832314809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,6144,0.017166222135225933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,8192,0.027642667293548584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,7168,0.013849777479966482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,7168,0.017837334010336135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,5120,0.015089778436554803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,7168,0.02436622149414486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,6144,0.016532444291644625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,6144,0.023147554861174688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,5120,0.014129777749379476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,4096,0.012740444805887012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,5120,0.011702222128709158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,5120,0.02015999952952067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,3584,0.010812444819344414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,10240,0.02676711148685879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,4096,0.012413333687517377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,4096,0.018806222412321303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,3584,0.01166133334239324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,3584,0.009630222287442949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,7168,0.019671110643280875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,3584,0.017162667380438913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,2560,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,65536,0.09137955639097427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,3072,0.010778666370444827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,3072,0.01625777781009674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,2560,0.010112000008424124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,2048,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,6144,0.012698666916953193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,2560,0.01569422251648373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,2560,0.008011555506123437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,1536,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,2048,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,2048,0.009309333231714036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,2048,0.014293332894643148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,4096,0.010620444185203977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,1536,0.0075839997993575195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,1536,0.006912888752089606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,1536,0.012578667038016848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,1024,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,768,0.004111999852789773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,3072,0.009935110807418823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,1024,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,1024,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,768,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,3072,0.008734222087595198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,768,0.006226666685607698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,256,0.002471999989615546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,768,0.010791111323568555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,512,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,512,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,128,0.0021262222694026097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,512,0.010067555639478896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,256,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,256,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,64,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,256,0.009407111340098912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,128,0.0058782220714622075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,128,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,32,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,128,0.007683555285135905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,64,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,1024,0.0047182221379545005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,8192,32,0.005320888840489917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,51200,0.11622844802008735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,65536,0.08253422048356798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,16384,0.04067111015319824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,65536,0.06618399752510919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,8192,512,0.0028248888750871024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,51200,0.05351377858055962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,12288,0.031861331727769636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,16384,0.022209778428077698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,16384,0.042740444342295326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,65536,0.13075110647413465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,10240,0.02701777716477712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,51200,0.10472977823681301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,12288,0.019941333267423842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,12288,0.017792888813548617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,8192,0.0233297778500451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,10240,0.01570933394961887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,10240,0.017320889565679762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,65536,0.1494871113035414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,10240,0.030415998564826116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,8192,0.013781333135233985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,7168,0.017902221944597032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,8192,0.015076445208655464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,6144,0.015614221493403116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,8192,0.02626755502488878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,7168,0.01292888902955585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,51200,0.06316088967853122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,7168,0.023745778534147475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,5120,0.013436444103717804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,6144,0.015072888798183866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,6144,0.012053333222866058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,4096,0.011529777612951068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,16384,0.025772444076008264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,6144,0.022202667262819078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,5120,0.010951999988820819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,3584,0.010056888891590966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,5120,0.013777777552604675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,4096,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,5120,0.020213334096802604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,4096,0.01202044470442666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,3072,0.009361777868535783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,3584,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,4096,0.01863911085658603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,12288,0.03484800126817491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,2560,0.008303110798199972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,3584,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,3584,0.016972444123691983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,3072,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,2048,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,3072,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,2560,0.007709333466158972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,3072,0.016163556112183463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,1536,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,2048,0.007249777515729268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,2048,0.008056888977686564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,1024,0.00452622233165635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,2048,0.013890667094124688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,1536,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,1536,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,768,0.003642666671011183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,1536,0.012404444317022959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,1024,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,1024,0.010715555813577441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,768,0.006001777946949005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,512,0.002838222309947014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,768,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,768,0.010709332923094431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,256,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,512,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,512,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,512,0.009179555707507664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,128,0.0021537778278191886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,256,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,256,0.008956444760163626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,2560,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,64,0.0018648888087934917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,2560,0.01571111049917009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,128,0.004951111144489712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,7168,32,0.0018906665758954154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,7168,0.016937777400016785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,128,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,65536,0.12115732828776042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,64,0.005404444618357553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,32,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,51200,0.09569155507617527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,65536,0.07597511344485812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,65536,0.06337777773539226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,16384,0.03444177905718485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,51200,0.05147200160556369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,7168,256,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,65536,0.1306462287902832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,16384,0.023511999183230933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,12288,0.027455111344655354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,51200,0.10541511244244045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,16384,0.02093511157565647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,16384,0.04199466771549649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,12288,0.017159110969967313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,12288,0.018178666631380718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,12288,0.034520887666278414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,8192,0.02025422288311852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,10240,0.015982222225930955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,10240,0.015230221880806817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,10240,0.03005689051416185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,8192,0.01371911085314221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,7168,0.01773866679933336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,8192,0.014855111638704935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,51200,0.05667644739151001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,6144,0.015835555063353646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,8192,0.025847999585999384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,7168,0.012434666355450949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,7168,0.017322666115230985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,6144,0.011704000333944956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,7168,0.023807111713621352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,6144,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,4096,0.01088177743885252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,6144,0.021747555997636583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,5120,0.010399111443095738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,5120,0.013070222404268054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,3584,0.009383999639087254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,4096,0.009249777429633671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,4096,0.011400000088744693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,4096,0.01788444485929277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,3584,0.008347555167145198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,3584,0.011076444553004371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,3584,0.017342221405771043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,3072,0.008273777862389883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,2560,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,3072,0.01644711030854119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,5120,0.013896889156765409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,2560,0.008993777963850234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,2048,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,2560,0.007352888584136963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,2560,0.01514311134815216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,128,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,5120,0.019868445065286424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,1536,0.0058764442801475525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,2048,0.007828444242477417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,2048,0.013519110778967539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,1024,0.0042204442951414315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,1536,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,3072,0.00850311087237464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,1536,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,768,0.003491555651028951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,1536,0.011370666325092316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,1024,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,1024,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,3072,0.010040889183680216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,512,0.00282222218811512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,1024,0.010770666930410596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,768,0.005976000179847081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,768,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,768,0.009975999593734741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,512,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,512,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,128,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,512,0.009296889106432596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,256,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,256,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,64,0.0018737777653667664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,256,0.008617777791288164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,128,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,32,0.0018933332628673976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,10240,0.023408000667889912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,2048,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,128,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,64,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,65536,0.1067884431944953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,32,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,51200,0.084279113345676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,65536,0.06730044550365873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,51200,0.05120088987880283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,16384,0.03094844354523553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,51200,0.04965955681271023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,65536,0.13036355707380506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,51200,0.10480799939897324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,16384,0.020280000236299302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,12288,0.02495288848876953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,16384,0.041704889800813466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,6144,128,0.004958222309748332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,10240,0.021172444025675457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,12288,0.016501333978441026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,12288,0.03382755650414361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,10240,0.014817777607176038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,10240,0.014879999889267815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,65536,0.06117244561513265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,8192,0.018422222799725003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,10240,0.029203553994496662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,7168,0.016200888488027785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,8192,0.013205333716339536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,6144,256,0.002303999952144093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,8192,0.025479111406538222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,6144,0.014126222994592456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,7168,0.01683999929163191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,16384,0.021374222305085924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,7168,0.023344889283180237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,5120,0.012336889074908363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,6144,0.014305777019924588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,6144,0.02169422143035465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,12288,0.01661422186427646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,4096,0.009687110781669617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,5120,0.012774222426944308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,5120,0.009922666682137383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,5120,0.01979200045267741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,3584,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,4096,0.008343111309740279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,4096,0.017551110850440133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,3072,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,3584,0.00793688909875022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,8192,0.013970666461520724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,3072,0.00979644474056032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,7168,0.012322666744391123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,2560,0.006700444552633498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,3072,0.015104000767072042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,6144,0.011066666907734342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,2560,0.007257777783605788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,2560,0.00831911133395301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,2560,0.014599111345079211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,2048,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,2048,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,1536,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,2048,0.0127697777416971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,4096,0.011390222443474663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,1024,0.0037084445357322693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,1536,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,1536,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,3584,0.010399111443095738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,1536,0.011401777466138205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,3584,0.017196445001496207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,1024,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,1024,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,512,0.00286666676402092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,1024,0.01016266644001007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,3072,0.007663110891977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,768,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,768,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,256,0.0022462221483389535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,512,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,512,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,128,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,2048,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,512,0.009336888790130615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,256,0.004972444640265571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,64,0.0018133332745896445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,128,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,32,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,128,0.004960000101062986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,128,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,64,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,65536,0.0850177804629008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,5120,32,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,5120,768,0.0032391111469931076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,51200,0.06654133399327596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,65536,0.060080885887145996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,65536,0.059325330787234835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,16384,0.025384889708624944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,51200,0.047694222794638745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,51200,0.04728622237841288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,65536,0.1301199992497762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,12288,0.019440889358520508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,768,0.010099555883142684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,16384,0.018977777825461496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,16384,0.019502222537994385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,51200,0.10443822542826335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,256,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,10240,0.016876444220542908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,256,0.008729777402347988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,16384,0.040744000011020236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,12288,0.01612622208065457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,8192,0.014831110835075378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,12288,0.033131556378470525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,10240,0.014304889572991265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,10240,0.0138568886452251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,8192,0.012749333348539142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,10240,0.028844445943832397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,8192,0.012839111189047495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,6144,0.011844444606039258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,8192,0.025477333201302424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,7168,0.011561777856614856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,7168,0.014038221703635322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,5120,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,7168,0.023397333092159692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,6144,0.010500444306267632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,6144,0.02130399975511763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,4096,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,5120,0.011959999799728394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,5120,0.01978577839003669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,4096,0.011426666544543372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,4096,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,3584,0.010994666980372535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,12288,0.01536622146765391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,4096,0.01812622282240126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,3072,0.009291555318567488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,3584,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,7168,0.01311555587583118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,3584,0.00998222248421775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,2560,0.00654666663871871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,3072,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,3072,0.007625777688291337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,3072,0.015467555986510383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,2560,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,2048,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,2560,0.007407111426194509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,6144,0.01277955538696713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,1536,0.003963555726740095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,2560,0.014081777797804939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,2048,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,5120,0.008930666579140557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,2048,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,1024,0.0032248888164758682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,2048,0.012418666647540199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,1536,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,1536,0.006311111152172089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,768,0.0028080000645584534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,1536,0.011766222616036734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,1024,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,1024,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,1024,0.010348444183667501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,768,0.005984000033802456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,768,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,256,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,768,0.01032266683048672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,512,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,512,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,128,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,3584,0.016134222348531086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,512,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,256,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,64,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,256,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,128,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,256,0.00905600024594201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,65536,0.07820977767308553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,128,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,128,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,64,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,4096,32,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,65536,0.0574515528149075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,65536,0.05823377768198649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,16384,0.02236000034544203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,51200,0.04673599865701464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,51200,0.04519911275969612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,12288,0.017998221847746108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,65536,0.1308844486872355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,4096,512,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,51200,0.10463111268149482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,16384,0.018290667070282828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,10240,0.015592000550693937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,16384,0.04058577617009481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,12288,0.015849777393870883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,12288,0.014858666393491956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,8192,0.013391999734772576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,10240,0.014207111464606391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,12288,0.03268444538116455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,7168,0.012087110843923358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,10240,0.02922044528855218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,8192,0.012191111014948951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,8192,0.0124195557501581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,6144,0.011231111155615913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,8192,0.025041777226659987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,7168,0.011015999648306104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,7168,0.013406222065289816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,51200,0.05939110782411364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,7168,0.023709333605236475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,5120,0.010015111002657149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,6144,0.009658666948477427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,6144,0.01240711079703437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,6144,0.021301334102948506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,4096,0.010832000109884473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,5120,0.009019555317031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,5120,0.01163999984661738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,3584,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,5120,0.02014400064945221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,16384,0.01924088928434584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,4096,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,4096,0.009852444132169088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,3072,0.008990222381220924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,4096,0.01720533271630605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,3584,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,2560,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,3584,0.008646222452322641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,3072,0.007358222372002072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,3584,0.016612443659040663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,3072,0.00793688909875022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,2048,0.00489688871635331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,3072,0.01482311056719886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,2560,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,2560,0.007223111059930589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,1536,0.003912000192536248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,2048,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,2560,0.013944889108339945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,2048,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,1024,0.0031742221779293488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,2048,0.012758221891191272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,1536,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,768,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,1536,0.011331555744012197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,1024,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,1024,0.005958222266700532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,512,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,768,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,768,0.005884444548024072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,768,0.00997511131895913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,256,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,512,0.005877333382765452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,512,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,512,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,256,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,128,0.0021671110557185277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,10240,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,256,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,64,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,128,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,128,0.005013333426581489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3584,32,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,128,0.0073626670572492815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,65536,0.06136177645789253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,64,0.005192000004980299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,1536,0.006080889039569431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,32,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,51200,0.049118220806121826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,65536,0.05536444319619072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,1024,0.010372444159454769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,16384,0.019323555959595572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,65536,0.05747377872467041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,51200,0.045913779073291354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,51200,0.04361244373851352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,65536,0.13034221861097547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,12288,0.015817777978049386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,16384,0.01889866590499878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,10240,0.013771555489963956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,16384,0.018231999542978074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,51200,0.10413599676556057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,16384,0.04016710983382331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,12288,0.015678221980730694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,8192,0.011969777444998423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,12288,0.01443911095460256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,10240,0.013765333427323235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3584,256,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,7168,0.011074666347768573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,10240,0.013056889176368713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,10240,0.02846222122510274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,8192,0.011594666375054253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,8192,0.01184088902340995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,6144,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,7168,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,8192,0.02529777752028571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,5120,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,7168,0.013167999684810638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,6144,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,7168,0.02312533391846551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,4096,0.00869688888390859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,6144,0.011239999698268043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,5120,0.008656889200210571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,6144,0.021319111188252766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,3584,0.00941066692272822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,4096,0.008230222596062554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,5120,0.01956888867749108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,3072,0.008504889077610439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,4096,0.016537777251667447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,3584,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,3584,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,3584,0.015820443630218506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,2560,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,3072,0.007320000065697565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,3072,0.014259555273585849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,12288,0.03258488906754388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,2560,0.007328888608349695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,2560,0.013402666482660504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,2048,0.006654222392373615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,2048,0.012085333466529846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,1536,0.003830222205983268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,5120,0.01036888857682546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,2048,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,4096,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,1024,0.0031866667171319327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,1536,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,1536,0.011756444142924415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,3072,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,768,0.002805333377586471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,1024,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,1024,0.010327999790509542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,2560,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,512,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,768,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,768,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,768,0.01036177741156684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,256,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,512,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,512,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,512,0.009709333380063375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,128,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,256,0.004952888935804367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,256,0.008960888617568547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,1536,0.006656888872385025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,64,0.0018195555441909367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,128,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,32,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,128,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,1024,0.005852444304360284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,65536,0.05640711386998495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,64,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,3072,32,0.004960888789759742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,51200,0.043473776843812734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,3072,2048,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,65536,0.05641422006818983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,65536,0.05332444442643059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,16384,0.017828444639841717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,65536,0.13027200433943006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,51200,0.0451822214656406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,51200,0.04172711239920723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,12288,0.014110222458839417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,16384,0.017154667112562392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,16384,0.018876444962289598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,10240,0.012224888636006249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,16384,0.03974044322967529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,12288,0.015466666883892484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,12288,0.014381332529915703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,8192,0.010704889065689512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,10240,0.013066666821638743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,12288,0.03223022156291538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,128,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,7168,0.009848888549539778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,10240,0.02907555633121067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,8192,0.011027555498811932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,8192,0.011520889070298938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,6144,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,7168,0.010411555568377176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,8192,0.02513155506716834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,5120,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,7168,0.02299822204642826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,6144,0.009684444301658207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,6144,0.009988444546858469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,6144,0.021293333835071985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,5120,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,5120,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,5120,0.018908444378111098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,3584,0.008349333372380998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,4096,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,51200,0.10432444678412543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,4096,0.01684266659948561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,3072,0.00758844448460473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,3584,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,10240,0.012717333104875354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,3584,0.01551466683546702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,2560,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,3072,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,3072,0.014463111758232117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,7168,0.011512888802422417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,2048,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,2560,0.0070168889231152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,2560,0.013780444032616086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,4096,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,1536,0.0038542221817705366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,2048,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,2048,0.006984000404675801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,2048,0.012433778080675336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,4096,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,1536,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,1536,0.005881777654091517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,1536,0.011362666885058085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,3584,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,1024,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,1024,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,1024,0.010739555789364709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,3072,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,768,0.005967999911970562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,512,0.00254133323000537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,768,0.005408888889683618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,768,0.01034222212102678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,2560,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,256,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,512,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,512,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,512,0.009298666483826106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,256,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,256,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,256,0.009361777868535783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,64,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,128,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,128,0.004853333450025982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,1024,0.003168888803985384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,128,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,65536,0.04867022236188253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,64,0.004948444250557157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2560,32,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,768,0.002846222163902389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,51200,0.035368889570236206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,65536,0.05088444550832113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,65536,0.05550933215353224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,16384,0.014382221632533602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,65536,0.13007377253638372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,51200,0.0446320010556115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,51200,0.04059200154410468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,12288,0.011731555892361535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,16384,0.018231110440360177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,16384,0.016843555702103507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,128,0.001875555556681421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,10240,0.010842666857772403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,16384,0.03975555631849501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,12288,0.013795555465751223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,12288,0.014161777165200977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,8192,0.009314667019579146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,12288,0.03224888775083754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2560,32,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,10240,0.01275288893116845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,7168,0.008850666383902231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,10240,0.028796444336573284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,8192,0.010671999719407825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,8192,0.010850666297806634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,6144,0.011236444115638733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,8192,0.025016888976097107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,7168,0.01033866653839747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,7168,0.010009778042634329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,5120,0.010018666585286459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,7168,0.022653333014912073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,6144,0.009309333231714036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,6144,0.009089777866999308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,5120,0.00867377801073922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,5120,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,5120,0.01830933325820499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,51200,0.10373155275980632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,3584,0.008016889293988546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,4096,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,4096,0.008310221963458592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,3072,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,4096,0.016186666157510545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,3584,0.007663110891977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,3584,0.007336888876226213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,2560,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,10240,0.012276444170210095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,3072,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,3584,0.01564711166752709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,3072,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,2048,0.0042071110672420925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,2560,0.0069982219073507525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,3072,0.014465777410401238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,1536,0.003863111138343811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,2560,0.013170666992664337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,2048,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,2048,0.012721777790122561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,1024,0.0031715554909573663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,1536,0.006404444575309753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,1536,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,4096,0.008984000318580205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,768,0.0028488888508743713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,6144,0.02130488885773553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,1024,0.006206222292449739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,1024,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,512,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,1024,0.010061333576838175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,768,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,768,0.005459555735190709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,256,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,768,0.010051555931568146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,512,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,512,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,128,0.0019315555691719055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,512,0.009311999711725447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,256,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,256,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,2560,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,64,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,2048,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,128,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,128,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,2048,32,0.001818666648533609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,65536,0.034881777233547635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,64,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,32,0.005014222115278244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,1536,0.011416888899273343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,65536,0.049119111564424306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,51200,0.027404444085227117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,65536,0.05467289023929172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,16384,0.012451555165979596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,65536,0.1299991077846951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,51200,0.044340444935692676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,51200,0.039332442813449435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,12288,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,16384,0.017383999294704862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,51200,0.1031457781791687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,16384,0.016428444120619033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,256,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,10240,0.00889866633547677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,16384,0.03972444600529141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,12288,0.013746666411558786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,12288,0.013455111119482251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,8192,0.007909333540333642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,2048,128,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,10240,0.012096888489193387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,12288,0.03239466746648153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,7168,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,10240,0.028488887680901423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,8192,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,8192,0.009842666486899057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,6144,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,8192,0.024326221810446844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,7168,0.010001777774757808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,7168,0.009630222287442949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,5120,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,7168,0.022311111291249592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,6144,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,4096,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,5120,0.009005332986513773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,6144,0.020248888267411124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,5120,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,3584,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,5120,0.018219555417696636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,4096,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,4096,0.00830844458606508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,3072,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,4096,0.016357334123717416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,3584,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,3584,0.007334222396214803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,2560,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,3584,0.01516711049609714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,3072,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,3072,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,2048,0.00572533326016532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,10240,0.011733333269755045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,2560,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,3072,0.014555555250909595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,2560,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,1536,0.003546666767862108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,2560,0.013129777378506131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,1024,0.0031617778456873367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,2048,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,2048,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,1536,0.006304889089531369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,768,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,2048,0.012182222472296821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,1536,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,1024,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,6144,0.008905777500735389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,1536,0.011798222031858234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,1024,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,768,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,1024,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,256,0.002128888956374592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,768,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,512,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,768,0.009696000152164036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,128,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,512,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,512,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,64,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,256,0.004945777770545748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,128,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,32,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,256,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,65536,0.024497777223587036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,128,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,128,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,51200,0.019734222027990553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,64,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1536,32,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,65536,0.0544088880221049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,65536,0.04692444536421034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,16384,0.013148444394270578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,51200,0.04375911090109083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,65536,0.12953155570560032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,51200,0.03805066810713874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,12288,0.011215999722480774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,16384,0.015662221444977652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1536,512,0.002529777793420686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,16384,0.039048890272776283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,16384,0.015863999724388123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,10240,0.010017777482668558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,12288,0.012392000191741519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,12288,0.013419555293189155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,8192,0.009215110705958473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,12288,0.031594667169782854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,10240,0.011081777513027191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,10240,0.011729777687125735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,256,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,7168,0.00869333330127928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,10240,0.027850667635599773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,8192,0.01074488874938753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,8192,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,6144,0.007786666353543599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,7168,0.010045333041085137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,8192,0.023983110984166462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,7168,0.009319110876984065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,5120,0.007256000406212277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,7168,0.022314666046036616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,6144,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,4096,0.005987555616431766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,6144,0.008719111482302347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,6144,0.020287111401557922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,5120,0.008065777520338694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,3584,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,5120,0.01809244520134396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,4096,0.008238222036096785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,4096,0.007490666376219855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,51200,0.10375377866956924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,3072,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,4096,0.01590488851070404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,3584,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,3584,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,3072,0.007335999773608313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,3584,0.015492444237073263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,3072,0.0070062221752272705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,2048,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,3072,0.014160000615649752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,2560,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,2560,0.012805333567990197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,1536,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,2048,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,2048,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,2048,0.012061333490742577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,1024,0.003914666672547658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,1536,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,1536,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,5120,0.008607111043400234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,1536,0.011357333097192975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,768,0.0032675556010670135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,1024,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,1024,0.010777778095669217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,512,0.0026311110705137253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,768,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,768,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,2560,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,768,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,256,0.0022008888837363985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,512,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,512,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,256,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,2560,0.0063688887490166565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,256,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,64,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,256,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,32,0.002116444417172008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,128,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,128,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,65536,0.01960177719593048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,128,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,64,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,51200,0.016590222716331482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,1024,32,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,65536,0.046273777882258095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,1024,0.0057813334796163775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,16384,0.013400889105266996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,65536,0.12909777959187826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,51200,0.043471111191643604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,51200,0.10234222147199844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,16384,0.01585422290696038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,12288,0.010052444206343757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,16384,0.014845333165592618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,16384,0.03903377718395657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,1024,128,0.002128888956374592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,10240,0.008290666672918532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,512,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,12288,0.01312355531586541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,8192,0.00720088928937912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,12288,0.03119200136926439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,10240,0.011744000017642975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,10240,0.011213333242469363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,7168,0.007376888559924231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,10240,0.02738933265209198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,8192,0.009712888962692684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,8192,0.02338933282428318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,65536,0.05396622088220385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,7168,0.009331555830107795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,6144,0.006649777707126405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,7168,0.009863999982674917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,7168,0.02160266704029507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,5120,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,51200,0.037307557132509016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,6144,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,6144,0.008747555315494537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,4096,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,6144,0.019927110936906602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,5120,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,5120,0.017877333694034152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,3584,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,4096,0.00796977761718962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,4096,0.007369777394665613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,12288,0.012311999996503195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,3072,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,4096,0.016175111134847004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,3584,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,3584,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,2560,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,8192,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,3584,0.01514488955338796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,3072,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,2048,0.004230222354332606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,3072,0.014131555954615274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,2560,0.006728000111050076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,2560,0.006461333483457565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,1536,0.004505777938498391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,2560,0.012921777864297232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,2048,0.006653333289755716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,2048,0.012710221939616732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,1536,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,1536,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,5120,0.008093333078755273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,1536,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,768,0.0029760001020299066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,1024,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,1024,0.00573333352804184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,1024,0.010058666268984476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,512,0.002537777854336633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,768,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,768,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,768,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,512,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,256,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,3072,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,512,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,128,0.0020702222569121257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,512,0.009326222042242685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,256,0.005633777628342311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,256,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,256,0.008647999829716152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,32,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,128,0.004856889032655292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,65536,0.027090667022599116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,128,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,768,1024,0.003504000190231535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,128,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,51200,0.02216533323129018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,64,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,32,0.004820444517665439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,65536,0.05319199959437052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,16384,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,65536,0.045838223563300244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,51200,0.038366221719317965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,65536,0.12823732693990073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,51200,0.03696533375316196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,16384,0.015571556157535978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,16384,0.014444443914625378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,51200,0.10324888759189182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,10240,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,16384,0.038358221451441445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,12288,0.013058666553762225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,8192,0.006228444476922353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,12288,0.030812445614073012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,10240,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,10240,0.027115555273161993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,7168,0.0052871112194326185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,8192,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,8192,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,768,2048,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,6144,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,8192,0.02334755493534936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,7168,0.009696889254781935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,7168,0.009317333499590555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,5120,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,7168,0.02163644466135237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,6144,0.009287999735938178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,6144,0.008683555656009251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,4096,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,6144,0.019905777441130746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,5120,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,5120,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,12288,0.008007999923494127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,3584,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,4096,0.00792711145348019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,4096,0.007340444458855524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,4096,0.015840889679061044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,12288,0.012255111502276527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,3584,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,3584,0.007711110843552484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,10240,0.011715555356608497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,3584,0.01483022173245748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,2560,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,3072,0.007627555893527136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,3072,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,2048,0.004593777573770947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,3072,0.014226666755146451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,2560,0.00647555540005366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,2560,0.01276177747382058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,1536,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,2048,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,2048,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,2048,0.012744888663291931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,1024,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,1536,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,1536,0.011083555718262991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,5120,0.018237334158685472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,768,0.00253866674999396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,1024,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,1024,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,3072,0.005205333232879639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,512,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,768,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,768,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,768,0.010384000009960597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,256,0.002209777840309673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,512,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,512,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,2560,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,128,0.0018595556418100994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,512,0.009704888694816166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,256,0.005295111073387994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,256,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,64,0.0018195555441909367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,128,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,128,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,512,32,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,65536,0.016597333881590102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,1536,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,64,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,512,32,0.004581333448489507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,1024,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,65536,0.043059554364946157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,65536,0.0465937786632114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,16384,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,51200,0.0373582210805681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,51200,0.034449779325061373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,16384,0.014126222994592456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,51200,0.10264266861809625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,16384,0.015815110670195687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,12288,0.00552977787123786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,16384,0.03834222091568841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,10240,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,12288,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,12288,0.012208888928095499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,256,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,8192,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,12288,0.03088888857099745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,128,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,10240,0.012054222325483957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,7168,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,10240,0.026779555612140234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,8192,0.01072355525361167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,8192,0.009709333380063375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,6144,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,51200,0.01424888935354021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,8192,0.023352889551056757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,7168,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,7168,0.009371555513805812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,5120,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,65536,0.1279040045208401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,6144,0.009314667019579146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,7168,0.021310221817758348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,4096,0.004534222185611725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,6144,0.00885333369175593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,5120,0.008397333323955536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,5120,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,3584,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,4096,0.008023111356629265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,5120,0.018173333671357896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,4096,0.007040888898902469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,3072,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,4096,0.016159110599093966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,3584,0.0074133334888352295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,2560,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,3584,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,3584,0.014829332629839579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,3072,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,3072,0.006897777732875612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,2048,0.00453244439429707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,3072,0.014116444521480136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,2560,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,10240,0.011064888702498542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,1536,0.003527111063400904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,2048,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,2560,0.013459555804729462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,2048,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,1024,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,2048,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,1536,0.006502222269773483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,1536,0.010762666662534079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,768,0.0028444443725877335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,1024,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,1024,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,1024,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,6144,0.019778667224778067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,512,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,768,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,768,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,256,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,768,0.010397333237859938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,512,0.005036444299750858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,512,0.008962666822804345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,128,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,256,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,256,0.004602666530344221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,256,0.009327111144860586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,2560,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,64,0.0018008889423476325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,128,0.0052968888647026485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,256,32,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,128,0.004864888886610667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,65536,0.011578666667143503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,128,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,1536,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,32,0.004553777890072929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,51200,0.010012444522645738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,65536,0.04582577612664965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,65536,0.0421982208887736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,16384,0.005679111099905438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,65536,0.12724799580044216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,51200,0.0370399985048506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,51200,0.03393155667516921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,12288,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,51200,0.1015511088901096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,16384,0.014281777871979607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,512,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,10240,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,16384,0.038350221183564924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,12288,0.013070222404268054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,12288,0.011894221935007306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,8192,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,10240,0.011706666813956367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,12288,0.03118577930662367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,10240,0.01072177787621816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,7168,0.005176888985766305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,10240,0.027407111393080816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,8192,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,8192,0.009510222408506606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,6144,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,8192,0.02351377738846673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,256,64,0.0047493332790003884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,7168,0.009072889056470659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,5120,0.004536889079544279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,7168,0.02198133369286855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,6144,0.009333333207501305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,6144,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,4096,0.004574222283230888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,5120,0.008623110751310984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,5120,0.008233778178691864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,3584,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,5120,0.017497777938842773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,16384,0.015777778294351365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,4096,0.008049777812427944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,4096,0.006999111009968652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,3072,0.004583111239804162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,4096,0.015822221835454304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,3584,0.007322666545708974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,2560,0.004549333204825719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,3584,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,3072,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,3072,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,2048,0.0042257776690853965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,3072,0.01444533301724328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,2560,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,2560,0.006555555595291986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,1536,0.003495111233658261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,2560,0.013462222284740873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,2048,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,2048,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,7168,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,1024,0.0028435554769304064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,2048,0.012740444805887012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,1536,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,1536,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,768,0.002521777732504739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,1536,0.01167555567291048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,1024,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,1024,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,6144,0.019556444552209642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,768,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,768,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,256,0.0023831110447645187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,512,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,512,0.00903733323017756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,3584,0.015047111445003085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,256,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,128,0.0018257778137922287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,256,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,256,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,128,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,128,0.004619555754794015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,128,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,32,0.0017875555074877208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,64,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,512,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,65536,0.009118222528033787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,768,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,51200,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,32,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,16384,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,12288,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,128,512,0.0053075556125905775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,65536,0.04214577873547872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,10240,0.004675555560323927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,51200,0.034295111894607544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,16384,0.014422222971916199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,12288,0.011966221862369113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,10240,0.011038222246699862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,6144,0.0047662220895290375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,7168,0.00904177791542477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,5120,0.00425866660144594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,128,64,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,4096,0.003866666720973121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,6144,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,3584,0.004598222259018156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,5120,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,3072,0.004257777912749184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,4096,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,1024,0.01036711119943195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,3584,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,2048,0.004541333350870344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,3072,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,1536,0.0035377778112888336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,8192,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,1024,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,7168,0.004549333204825719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,768,0.002463111033042272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,512,0.0021840000732077495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,1536,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,8192,0.009711999860074785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,1024,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,128,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,768,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,64,0.0017946666727463405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,32,0.0018577778504954444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,256,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,128,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,64,0.004864888886610667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,32,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,51200,0.006709333509206772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,16384,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,64,2560,0.004516444272465176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,12288,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,65536,0.042352000872294106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,51200,0.034287999073664345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,16384,0.014323555760913424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,8192,0.004275555411974589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,2560,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,12288,0.011732444167137146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,7168,0.004584889031118817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,6144,0.004564444637960858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,10240,0.01106844428512785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,5120,0.004519999855094486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,8192,0.009675555759006077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,4096,0.0041759999261962045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,7168,0.009005332986513773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,3584,0.004214222232500712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,6144,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,512,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,3072,0.004247111164861255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,5120,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,2560,0.004208000169859992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,2048,0.0038942222793896994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,4096,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,1536,0.00350222239891688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,65536,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,3584,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,1024,0.002874666617976295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,3072,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,2560,0.006578666468461354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,512,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,2048,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,256,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,1536,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,128,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,1024,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,64,0.0017813332378864288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,768,0.005052444421582752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,10240,0.00508355556262864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,32,0.0018071112119489247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,512,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,256,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,128,0.004819555415047539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,64,0.004528000122971005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,32,32,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,16384,0.3226257695092095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,16384,0.1625422239303589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,51200,0.4892551104227702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,16,32,768,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,51200,0.962773323059082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,12288,0.2263031005859375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,12288,0.12204621897803412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,16384,0.1369831164677938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,16,64,2048,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,12288,0.10209421979056464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,10240,0.10292977756924099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,10240,0.1882995499504937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,10240,0.0885511106914944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,8192,0.15410666995578343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,7168,0.13268178039126927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,51200,0.4362613360087077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,6144,0.11463022232055664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,8192,0.11074399948120117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,8192,0.09472266832987468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,7168,0.06356710857815213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,7168,0.09340977668762207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,5120,0.096233778529697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,6144,0.0562302205297682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,4096,0.07919022109773424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,6144,0.08471377690633138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,5120,0.04682133264011807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,3584,0.06859999895095825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,5120,0.07020977470609877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,3072,0.05944355328877767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,4096,0.03976800044377645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,3584,0.035306665632459856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,2560,0.05019022358788384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,3584,0.0513991117477417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,3072,0.031481779283947416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,2048,0.04134933484925164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,3072,0.0452479985025194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,1536,0.031806220610936485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,2560,0.02736977736155192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,2560,0.03843555516666836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,1024,0.022329777479171753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,2048,0.027334221535258826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,2048,0.033430222007963396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,1536,0.01891377733813392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,1536,0.026522666215896606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,1024,0.015409777561823526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,512,0.012560000022252401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,256,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,1024,0.01995022263791826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,768,0.013243555194801755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,128,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,512,0.012675555215941535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,256,0.010444444086816577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,512,0.012640888492266336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,64,0.004252444538805219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,256,0.010010666317409938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,32,0.0042133331298828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,768,0.01611555616060893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,128,0.008016889293988546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,64,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,4096,0.06500266657935248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,65536,32,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,65536,768,0.017636444833543565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,65536,0.5155653423733181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,51200,0.3912968900468614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,16384,0.2777635521358914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,51200,0.8534444702996148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,65536,1.089498625861274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,16384,0.13095643785264757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,65536,128,0.010556444525718689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,12288,0.2087306711408827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,16384,0.11129511727227105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,10240,0.1750533315870497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,51200,0.3486631181504991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,65536,0.4376195536719428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,8192,0.14015200403001574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,12288,0.09846222400665283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,12288,0.08500444226794773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,7168,0.12199910481770833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,8192,0.08743377526601155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,10240,0.08322933647367689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,8192,0.07714400026533338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,6144,0.10545511378182305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,7168,0.05286133289337158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,5120,0.08860888746049668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,10240,0.06980711221694946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,4096,0.0715528859032525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,6144,0.06891466511620416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,6144,0.051500442955229014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,3584,0.06301244762208727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,5120,0.05769955449634128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,4096,0.03286133209864298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,3072,0.05544177691141764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,4096,0.052854223383797534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,3584,0.028926223516464233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,3584,0.04054488738377889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,3072,0.026039999392297532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,3072,0.03773244553142124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,2560,0.022428444690174524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,2560,0.03249511122703552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,1536,0.029000888268152874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,2048,0.01945955554644267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,2048,0.028600000672870215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,1024,0.020709334148301017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,1536,0.016467556357383728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,1536,0.022005332840813532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,768,0.015139554937680563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,1024,0.013167999684810638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,7168,0.07437066899405585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,1024,0.01649688846535153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,512,0.010676444404655032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,768,0.01201777739657296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,768,0.013997332917319404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,256,0.006497777584526274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,512,0.010692444112565784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,5120,0.03880711065398322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,128,0.003305777700410949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,512,0.0110880004035102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,64,0.002554666664865282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,256,0.009391110804345872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,32,0.0028817777832349143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,2560,0.046239998605516225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,256,0.008366222182909647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,51200,2048,0.03721422288152907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,128,0.007079111205206976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,64,0.00904888908068339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,51200,32,0.00885333369175593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,65536,0.31964267624749076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,51200,0.24963556395636666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,65536,0.1778106689453125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,16384,0.08971911006503636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,51200,0.1313528882132636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,65536,0.1630906661351522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,16384,0.04215022259288364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,51200,0.12390844027201335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,16384,0.047601779301961265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,10240,0.05187199844254387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,12288,0.03341155582004123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,8192,0.04315377937422859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,12288,0.036736889017952815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,10240,0.02903111113442315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,7168,0.03712622324625651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,10240,0.03176711002985636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,8192,0.024352888266245525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,6144,0.033082667324278094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,8192,0.03388266762097677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,5120,0.028148445818159316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,7168,0.022414222359657288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,7168,0.030749334229363337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,4096,0.0240000006225374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,6144,0.028057777219348486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,5120,0.017687999539905123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,5120,0.023473777704768713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,3584,0.020187555087937247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,4096,0.015632889337009855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,4096,0.02163022259871165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,3584,0.014333332578341166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,2560,0.015091554986106025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,12288,0.06983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,3584,0.017291555802027386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,3072,0.013229333692126803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,2048,0.01238222254647149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,3072,0.015929778416951496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,1536,0.009716444545321995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,2048,0.012703999876976013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,2560,0.014519110321998596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,1024,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,2560,0.012162666353914471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,2048,0.01348533315791024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,768,0.005980444451173146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,1536,0.009128889275921715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,1536,0.011516444385051727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,512,0.00443733317984475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,1024,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,768,0.007255111303594377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,1024,0.00906133320596483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,256,0.0025315555847353404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,512,0.007041778001520369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,768,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,6144,0.019928000039524503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,512,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,256,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,64,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,32,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,256,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,16384,128,0.006607999818192587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,3072,0.017639999588330586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,128,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,64,0.006005333529578314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,16384,32,0.0063164445261160536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,65536,0.2526888847351074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,51200,0.1979315545823839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,16384,0.06548444430033366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,65536,0.14857155746883816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,12288,0.05085955394638909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,65536,0.1353271139992608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,51200,0.10643021927939521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,16384,0.037408000893063016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,10240,0.04318133327696058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,16384,0.04133244355519613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,51200,128,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,12288,0.029735111527972754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,12288,0.03205777870284186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,8192,0.03580355644226074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,7168,0.03141599893569946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,10240,0.025958221819665697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,10240,0.02767555581198798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,8192,0.02716000046994951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,6144,0.02755555510520935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,8192,0.030663109487957422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,7168,0.01994044416480594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,5120,0.0240666667620341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,7168,0.026542221506436665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,6144,0.018039110634062026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,4096,0.019176888797018263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,6144,0.024455111887719896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,3584,0.016615110966894362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,5120,0.020890666378868952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,4096,0.01442399952146742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,3072,0.014595554934607612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,4096,0.01847555571132236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,3584,0.01332266628742218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,2560,0.012245333029164208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,3584,0.016694222887357075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,51200,0.11300533347659641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,2048,0.010681777364677854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,3072,0.014656000667148165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,2560,0.011224889092975192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,1536,0.008418666819731394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,2560,0.01348622226052814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,2048,0.012322666744391123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,1024,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,2048,0.009722666607962714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,16384,128,0.002238222294383579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,1536,0.010544000400437249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,768,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,1024,0.007489778101444244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,512,0.0038613333470291565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,768,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,5120,0.016180444094869826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,768,0.007061333292060428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,256,0.0028355556229750314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,512,0.006806222101052602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,128,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,512,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,256,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,64,0.0021973333011070886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,128,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,256,0.0058515556156635284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,12288,32,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,128,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,3072,0.012409778104888068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,64,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,65536,0.21667822202046713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,32,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,65536,0.1392364501953125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,65536,0.1280453337563409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,16384,0.05834666887919108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,51200,0.10790489117304485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,51200,0.09950933191511367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,12288,1024,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,16384,0.035184890031814575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,16384,0.03976444403330485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,10240,0.037195556693606906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,8192,0.03127110997835795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,12288,0.030543999539481267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,10240,0.024715556038750544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,10240,0.026748443643252056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,7168,0.027832001447677612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,8192,0.024546666277779475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,8192,0.026923555466863845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,7168,0.019359111785888672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,7168,0.02194044490655263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,5120,0.021153777837753296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,6144,0.01755199995305803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,6144,0.02089866664674547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,51200,0.1721280018488566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,4096,0.018156444032986958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,5120,0.015846222639083862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,3584,0.015173332558737861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,5120,0.017631111873520743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,4096,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,3072,0.013447110851605734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,12288,0.04545244574546814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,3584,0.012887111140622033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,4096,0.01610311037964291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,3584,0.0144177774588267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,2560,0.011382222175598145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,2048,0.009862222605281407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,3072,0.011753777662913004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,12288,0.02755022214518653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,3072,0.013213333156373767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,1536,0.008040000167157915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,2048,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,2560,0.011757333245542316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,1024,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,1536,0.008228444390826756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,2048,0.010827555424637265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,768,0.004855999930037392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,6144,0.02469066613250309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,1536,0.009535111486911774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,1024,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,512,0.003896000070704354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,1024,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,768,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,256,0.002875555513633622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,768,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,512,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,128,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,512,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,256,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,64,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,128,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,10240,32,0.0021671110557185277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,256,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,128,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,65536,0.17260177930196127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,64,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,10240,32,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,51200,0.1296773354212443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,65536,0.09006577730178833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,16384,0.04442844457096524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,65536,0.0695448915163676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,10240,2560,0.010453333457310995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,12288,0.03499999973509047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,51200,0.06943022542529635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,51200,0.05768711037105984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,10240,0.026718222432666357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,16384,0.02330400049686432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,16384,0.027432888746261597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,12288,0.019620445039537217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,8192,0.024510223004553053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,12288,0.021591999464564856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,7168,0.020844444632530212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,10240,0.01639111174477471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,10240,0.018910222583346896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,8192,0.014632888966136508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,7168,0.013781333135233985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,8192,0.020280889338917203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,5120,0.015563555889659457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,7168,0.017838221457269456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,6144,0.012695999609099494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,12288,1536,0.00832622249921163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,4096,0.013200889031092325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,3584,0.011414222419261932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,5120,0.011485333244005838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,4096,0.010481778118345471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,5120,0.014853333433469137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,3072,0.010477333433098264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,4096,0.013817778064144982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,3584,0.009591999981138442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,2560,0.008951110972298516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,3584,0.011924444801277585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,2048,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,3072,0.008722667064931657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,3072,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,1536,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,2560,0.010103999740547603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,2048,0.00742133292886946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,1024,0.00479288895924886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,2048,0.00923733330435223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,1536,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,768,0.0035439998739295532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,1536,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,1024,0.00656444455186526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,512,0.002944888960984018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,1024,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,6144,0.018067555295096505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,256,0.0024399999529123306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,768,0.006028444402747684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,128,0.0021840000732077495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,512,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,768,0.006183111005359226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,256,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,512,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,32,0.0021395554973019496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,6144,0.01609244445959727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,256,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,128,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,65536,0.14847200446658665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,64,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,8192,32,0.005330666485759947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,51200,0.11638132731119792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,65536,0.08313422070609199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,65536,0.06625333097245958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,51200,0.06325955523384942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,51200,0.05401777889993456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,12288,0.03179733289612664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,16384,0.02168888847033183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,10240,0.026797332697444495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,16384,0.025756445195939805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,12288,0.017645332548353408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,8192,0.02310933338271247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,10240,0.015855110353893705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,10240,0.01753866672515869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,7168,0.01958577831586202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,8192,64,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,8192,0.014093332820468478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,6144,0.01705333259370592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,128,0.005719999886221356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,7168,0.01296622223324246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,7168,0.0166320006052653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,5120,0.014577777849303352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,6144,0.012005333271291522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,6144,0.015459555718633862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,4096,0.012489777472284106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,16384,0.04048888882001241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,5120,0.01386666629049513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,3584,0.011071999867757162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,4096,0.009672889278994666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,4096,0.01238844460911221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,3072,0.009754666851626502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,3584,0.008728000024954477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,3584,0.011394666300879585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,12288,0.01991911066903008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,2560,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,3072,0.008001777860853408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,3072,0.009893333746327294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,2048,0.0073724447025193115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,2560,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,1536,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,2560,0.00998044427898195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,2048,0.007087111473083496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,1024,0.004497777670621872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,2048,0.008565333154466417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,1536,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,768,0.00349866681628757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,1536,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,1024,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,1024,0.006342222293217977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,8192,2560,0.008420444197124904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,768,0.005993777679072485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,256,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,768,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,512,0.006019555446174409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,128,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,256,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,512,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,64,0.001869333287080129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,256,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,32,0.0018684443914228016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,128,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,128,0.004974222017659081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,65536,0.12148088879055446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,64,0.00545155546731419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,32,0.0052968888647026485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,51200,0.096233778529697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,7168,8192,0.019588443968031142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,16384,0.034075554874208235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,65536,0.0744204454951816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,65536,0.06361599763234456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,51200,0.051727109485202365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,16384,0.020724443925751578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,7168,512,0.0028151110228565005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,16384,0.023395554886923894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,7168,5120,0.011011555790901184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,10240,0.02335111134582096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,12288,0.017306667235162523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,8192,0.019983111156357657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,12288,0.01809244520134396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,10240,0.015136000182893542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,7168,0.017789334058761597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,10240,0.016125332978036668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,8192,0.013434666726324292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,7168,0.012446222205956778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,8192,0.018265777164035372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,5120,0.013756444056828817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,7168,0.01575999955336253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,6144,0.011807999677128263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,4096,0.010820444259378644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,6144,0.01568000018596649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,5120,0.010442666709423065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,3584,0.009391110804345872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,4096,0.008840888738632202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,3072,0.009283555878533257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,4096,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,3584,0.011363555159833698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,2560,0.008049777812427944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,3072,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,12288,0.02734666731622484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,3072,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,2048,0.006923555499977536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,2560,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,1536,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,2048,0.00720711135201984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,2560,0.008472888833946651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,1024,0.003964444415436851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,2048,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,1536,0.006673777682913675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,768,0.0032924444725116095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,1024,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,1536,0.007072000039948358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,512,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,1024,0.006312888943486744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,768,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,256,0.002297777682542801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,768,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,512,0.0057884446448749965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,128,0.0021359999146726397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,5120,0.012810666528013019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,256,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,3584,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,256,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,64,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,6144,128,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,128,0.004967111266321606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,51200,0.05701333284378052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,64,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,65536,0.10572177833980984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,32,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,51200,0.0839795536465115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,65536,0.06757599777645536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,65536,0.06146133608288235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,16384,0.03089600139194065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,6144,0.015777778294351365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,51200,0.051346666283077665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,51200,0.0495777792400784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,12288,0.024850666522979736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,10240,0.020964443683624268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,16384,0.019983111156357657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,16384,0.02127644419670105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,12288,0.01666933298110962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,8192,0.018425777554512024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,12288,0.016520889268981088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,10240,0.014846222268210517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,7168,0.01609511176745097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,6144,512,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,8192,0.013106666505336761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,10240,0.014885332849290637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,8192,0.016862221890025668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,5120,0.012449777788586087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,6144,0.011296889020336999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,7168,0.014821334017647637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,4096,0.01076533314254549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,6144,0.013766222529941134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,5120,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,5120,0.01236177815331353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,3584,0.009406222237481011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,4096,0.008360889222886827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,3072,0.008649778034951951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,4096,0.011350221931934357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,3584,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,2560,0.007463999920421177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,3584,0.01070666644308302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,3072,0.007826666865083907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,2048,0.006541333264774746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,3072,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,2560,0.007376888559924231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,1536,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,2048,0.007032000356250339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,2048,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,1024,0.003731555408901639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,1536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,1536,0.006853333363930385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,7168,0.012127999630239276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,768,0.0032471112079090546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,1024,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,1024,0.005972444597217772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,768,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,256,0.0023351110931899813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,768,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,512,0.00544177782204416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,128,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,256,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,256,0.005053333524200651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,128,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,32,0.0018666666001081467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,128,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,65536,0.08576711018880208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,64,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,2560,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,51200,0.06652710835138957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,6144,0.014270222849316068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,65536,0.059086223443349205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,6144,32,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,16384,0.025292444560262892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,5120,512,0.0028684443483750024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,51200,0.04706844356324938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,12288,0.020285333196322124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,51200,0.047592888275782265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,16384,0.01960266629854838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,10240,0.017625777257813346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,5120,512,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,16384,0.019546666079097323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,8192,0.015139554937680563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,12288,0.015455999308162265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,10240,0.014498665928840637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,7168,0.013045333325862885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,10240,0.013777777552604675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,8192,0.012709333664841123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,6144,0.011384888655609555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,8192,0.016579555140601266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,5120,0.010021333065297868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,7168,0.011418666276666852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,6144,0.01036800030204985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,7168,0.014108444253603617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,4096,0.009340444372759925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,6144,0.012882666455374824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,5120,0.00870400004916721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,3584,0.00776711106300354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,5120,32,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,4096,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,3072,0.007144889069928064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,65536,0.05973066886266073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,3584,0.0079644446571668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,2560,0.006260444306664997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,3584,0.008641777767075432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,3072,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,2048,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,3072,0.007687999970383114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,2560,0.006994666324721442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,1536,0.004077333129114575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,12288,0.016217778126398723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,2048,0.006993778049945831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,1024,0.003232888877391815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,1536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,1536,0.006568000134494569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,768,0.0028728888266616394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,1024,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,1024,0.005965333431959152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,512,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,768,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,768,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,5120,0.012262221839692859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,4096,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,512,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,128,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,512,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,256,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,64,0.0018853334089120228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,4096,128,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,32,0.001965333397189776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,128,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,65536,0.07707644171184964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,64,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,32,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,2048,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,51200,0.059303998947143555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,65536,0.058189332485198975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,65536,0.05769866704940796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,16384,0.022226666410764057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,51200,0.04674222071965536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,51200,0.045165333482954234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,16384,0.01926399932967292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,12288,0.017823111679818895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,10240,0.015428443749745687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,16384,0.018530666828155518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,4096,256,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,12288,0.015826667348543804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,12288,0.014855111638704935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,10240,0.014328888720936246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,7168,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,10240,0.013412444127930535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,256,0.004970666848950916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,6144,0.010936000280910067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,8192,0.01238844460911221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,8192,0.015441776977645027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,7168,0.011040888726711273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,5120,0.009565333525339762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,4096,0.008831111093362173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,7168,0.01369511087735494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,6144,0.01034311122364468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,4096,2560,0.0075937774446275495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,5120,0.008983111215962304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,3584,0.0074746666683091064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,6144,0.013321778012646569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,3072,0.006911110960774952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,4096,0.008289777570300633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,5120,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,2560,0.005979555762476391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,3584,0.008030222521887885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,4096,0.009496000077989366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,2048,0.004957333207130432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,3072,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,3584,0.00869066682126787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,1536,0.0038853333228164246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,2560,0.00721777809990777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,3072,0.007584888901975419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,1024,0.0032071111102898917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,2560,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,2048,0.006984000404675801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,768,0.002809777855873108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,1536,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,2048,0.0069955554273393415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,512,0.002485333424475458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,1536,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,768,0.005965333431959152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,1024,0.005730666634109285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,256,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,8192,0.013423110875818463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,512,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,512,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,256,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,32,0.0018826667219400406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,128,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,256,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,65536,0.06191466914282905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,64,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,32,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,65536,0.055606219503614635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,65536,0.05726222197214762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,51200,0.04594399862819248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,51200,0.04363200068473816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,16384,0.019243554936514962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3584,1024,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,12288,0.01585422290696038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,16384,0.018979556030697294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3584,128,0.0019040000107553268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,768,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,10240,0.015120888749758402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,12288,0.015616888801256815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,8192,0.013425778183672162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,12288,0.014469333820872836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,10240,0.013740444348918067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,7168,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3584,128,0.004865777575307422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,10240,0.01313866674900055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,8192,0.011368888947698804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,6144,0.011033777561452655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,8192,0.014145778285132514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,7168,0.01035377797153261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,5120,0.009008000294367472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,6144,0.010076444182131026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,7168,0.01348444405529234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,4096,0.008168000313970778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,6144,0.01167377746767468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,5120,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,3584,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,4096,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,5120,0.010280888941552904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,3072,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,4096,0.0088755554623074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,3584,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,2560,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,3584,0.007644444704055786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,3072,0.007719111111429002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,2048,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,3072,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,2560,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,1536,0.003970666478077571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,2048,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,2560,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,1024,0.003156444471743372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,2048,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,768,0.0028479999552170434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,1536,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,1024,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,512,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,1024,0.00565155554148886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,768,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,256,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,768,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,512,0.005628444254398346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,128,0.002012444453106986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,512,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,256,0.005632888939645555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,64,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,256,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,128,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,65536,0.05675377448399862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,64,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,32,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,51200,0.043741332160102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,65536,0.05635999970965915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,65536,0.053202665514416166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,1536,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,16384,0.019681778219011094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,51200,0.041760888364579946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,51200,0.04520444406403435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,12288,0.016303999556435477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,16384,0.018909333480728995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,10240,0.014119999276267158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,16384,0.017095110482639737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,12288,0.015153777268197803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,8192,0.012715555727481842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,12288,0.014151111245155334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,51200,0.049201776583989464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,10240,0.012840888566441007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,7168,0.011374221907721626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,10240,0.01276266657643848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,3072,32,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,6144,0.01015911085738076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,8192,0.013406222065289816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,7168,0.01037688884470198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,5120,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,7168,0.011991110940774282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,6144,0.009709333380063375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,4096,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,6144,0.010112889111042023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,5120,0.008739555875460306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,3584,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,5120,0.008463111188676622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,3072,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,3072,16384,0.017664889494578045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,3584,0.008027555214034187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,2560,0.005314666777849197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,3584,0.0074879998962084455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,3072,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,3072,0.007027555671003129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,2048,0.004222222086456087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,2560,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,2560,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,1536,0.003871110992299186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,2048,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,2048,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,1024,0.0031848889258172777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,1536,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,3072,128,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,1536,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,768,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,1024,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,512,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,768,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,1024,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,256,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,768,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,512,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,128,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,4096,0.008007999923494127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,512,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,256,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,4096,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,32,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,128,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,65536,0.04878933231035868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,64,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,32,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,51200,0.03559111224280463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,65536,0.05544177691141764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,65536,0.05037066671583387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,16384,0.016170667277442086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2560,8192,0.011035555766688453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,51200,0.044857776827282377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,51200,0.040405333042144775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,12288,0.01332711097266939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,16384,0.01847022275129954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,16384,0.016674665941132438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,10240,0.01203555530971951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,12288,0.014615111880832247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,8192,0.011415999796655444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,12288,0.013803555733627744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,10240,0.012283555335468717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,7168,0.009696000152164036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2560,64,0.0018213333355055915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,10240,0.012455999851226807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,8192,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,256,0.004947555561860402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,6144,0.008976889153321585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,8192,0.01144088887506061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,5120,0.008099555969238281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,7168,0.0103502223889033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2560,128,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,6144,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,4096,0.007270221908887227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,7168,0.010699555277824402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,3584,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,5120,0.008647999829716152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,6144,0.008887999587588841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,3072,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,4096,0.008023999631404877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,5120,0.008373333348168267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,2560,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,3584,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,4096,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,3072,0.007660444411966536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,1536,0.0037120001183615792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,2560,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,3584,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,3072,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,1024,0.003140444349911478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,2048,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,768,0.002814222127199173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,2048,0.006254222244024277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,1024,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,1536,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,1024,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,768,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,256,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,768,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,512,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,128,0.0019342222561438878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,256,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,512,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,64,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,256,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,32,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,128,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,65536,0.034734222624037005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,64,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,32,0.004960888789759742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,2048,0.00423555572827657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,51200,0.027592000034120347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,65536,0.054746667544047035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,16384,0.012777778009573618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,65536,0.04898400108019511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,51200,0.04433600107828776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,2048,2560,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,12288,0.011343999869293638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,1536,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,51200,0.038776887787712946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,16384,0.016538666354285348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,10240,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,2048,512,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,16384,0.0162462227874332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,8192,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,12288,0.01348533315791024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,7168,0.008599110775523717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,12288,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,10240,0.012077333198653327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,8192,0.011021333436171213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,6144,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,10240,0.011694221860832639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,5120,0.007110222346252865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,8192,0.010258666343159145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,7168,0.010318222145239512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,4096,0.0069439998931354946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,7168,0.009718221922715506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,6144,0.009326222042242685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,2048,128,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,3584,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,5120,0.008783111141787635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,6144,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,3072,0.005522666705979242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,4096,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,2560,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,4096,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,3584,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,3072,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,2048,0.004209777961174647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,3072,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,2560,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,1536,0.003567111161020067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,2560,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,1024,0.0031884445084465873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,2048,0.006034666465388403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,1536,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,768,0.002805333377586471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,1024,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,1024,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,512,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,768,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,768,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,256,0.0021359999146726397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,512,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,512,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,128,0.0018853334089120228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,256,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,5120,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,64,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,128,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,128,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1536,32,0.0018097776919603348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,65536,0.028102222416136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,2048,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,64,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,1536,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,51200,0.021984888447655573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,32,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,16384,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,65536,0.046932445632086865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,51200,0.04376888937420315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,16384,0.015997333659066092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,12288,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,51200,0.03777777817514207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,10240,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,16384,0.015557333827018738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,12288,0.01312800000111262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,8192,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,10240,0.012117333710193634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,12288,0.012440889245933957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,7168,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,10240,0.011072888970375061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,8192,0.01069866700304879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1536,256,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,8192,0.009983110758993361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,5120,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,7168,0.009371555513805812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,6144,0.00963644435008367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,4096,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,6144,0.00870400004916721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,5120,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,3584,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,5120,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,4096,0.008039111064540016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,3072,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,4096,0.007592000067234039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,3584,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,2560,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,3584,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,3072,0.007644444704055786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,2048,0.005719999886221356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,3072,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,2560,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,1536,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,2560,0.006562666760550604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,2048,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,1024,0.003875555677546395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,1536,0.006304889089531369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,6144,0.007553777760929531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1536,3584,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,7168,0.009846222069528367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,1024,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,1024,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,512,0.002817777709828483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,768,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,65536,0.05417422122425503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,512,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,512,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,256,0.002151111140847206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,128,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,256,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,256,0.004948444250557157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,1024,128,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,128,0.004860444615284602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,32,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,2048,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,65536,0.021868444151348535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,64,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,1536,0.005700444595681296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,32,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,51200,0.017539555827776592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,65536,0.054014222489462964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,65536,0.04640355706214905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,16384,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,51200,0.0428017775217692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,12288,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,1024,768,0.0053119998839166425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,16384,0.015807999504937064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,10240,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,12288,0.013063111239009433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,12288,0.012136889000733694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,8192,0.005977777971161737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,10240,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,10240,0.011030221978823343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,7168,0.006812444577614467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,64,0.0018115554832749898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,8192,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,6144,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,8192,0.009739555418491364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,7168,0.010064889159467485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,5120,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,6144,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,6144,0.00869333330127928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,1024,768,0.0032257777121331957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,4096,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,5120,0.008661333057615492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,3584,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,5120,0.008263111114501953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,4096,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,3072,0.0052871112194326185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,3584,0.007696888513035244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,4096,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,51200,0.03733511103524102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,2560,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,3584,0.007224889265166388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,3072,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,16384,0.015042665931913586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,1536,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,2048,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,2560,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,1024,0.003543111185232798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,2048,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,1536,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,768,0.0031519999934567344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,1536,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,1024,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,512,0.002624888800912433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,1024,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,768,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,256,0.0021768889079491296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,7168,0.009343110852771336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,768,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,512,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,128,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,512,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,256,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,32,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,128,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,256,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,65536,0.01591199967596266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,128,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,51200,0.013156444662147097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,64,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,768,2048,0.004221333397759331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,32,0.004861333303981357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,16384,0.007673777639865875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,768,2560,0.0069004446268081665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,51200,0.03870577613512675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,12288,0.006709333509206772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,51200,0.03666222095489502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,16384,0.01583377851380242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,10240,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,16384,0.01443288889196184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,12288,0.013056889176368713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,8192,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,12288,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,10240,0.011772444678677453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,7168,0.0053386667536364655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,8192,0.010684444672531553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,10240,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,6144,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,8192,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,7168,0.009655111365848118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,5120,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,6144,0.009304888546466827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,7168,0.009211555951171452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,4096,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,6144,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,5120,0.008393777741326226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,3584,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,5120,0.007916444705592262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,4096,0.008051555189821456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,3072,0.0053119998839166425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,4096,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,65536,0.04842311143875122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,3584,0.007631111476156447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,65536,0.04582755433188545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,768,3072,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,3584,0.0069013333155049225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,3072,0.007247111035717859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,2048,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,3072,0.006679111056857639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,1536,0.003543111185232798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,2048,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,2560,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,1024,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,1536,0.006310222049554189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,2048,0.005971555494599872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,768,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,1536,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,1024,0.005983110931184556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,512,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,768,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,256,0.0021520000365045336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,768,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,512,0.00564533347884814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,128,0.0019075555933846368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,512,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,256,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,64,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,256,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,128,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,128,0.004960000101062986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,65536,0.010694222317801582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,64,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,2560,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,51200,0.00942488925324546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,32,0.004605333424276776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,65536,0.046534223688973315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,512,2560,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,51200,0.037321776151657104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,12288,0.00518133325709237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,51200,0.03480266531308492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,16384,0.014471999473041959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,10240,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,12288,0.01329066687160068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,12288,0.012058667010731168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,8192,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,10240,0.011983111500740051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,10240,0.011139555937714048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,7168,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,8192,0.010520888699425591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,8192,0.009641778137948778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,512,32,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,7168,0.010227555202113258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,7168,0.009328000247478485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,5120,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,6144,0.009312888814343346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,6144,0.008642666869693333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,4096,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,65536,0.042479998535580106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,5120,0.008660444782839881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,5120,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,3584,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,16384,0.015822221835454304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,4096,0.008010666403505538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,4096,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,3072,0.004574222283230888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,3584,0.007725333174069722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,2560,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,3584,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,512,1024,0.0052844443255000645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,2048,0.00462400002612008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,3072,0.006788444601827198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,1536,0.0035333335399627686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,2560,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,2048,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,1024,0.0028942223224374983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,2048,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,1536,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,6144,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,768,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,1536,0.005697777701748743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,1024,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,1024,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,768,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,768,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,16384,0.0052986666560173035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,512,0.005655999812814925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,512,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,128,0.0018604445374674266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,256,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,256,0.004583111239804162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,64,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,128,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,128,0.004548444516128964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,32,0.001812444378932317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,65536,0.007793777518802219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,3072,0.007268444531493717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,256,2560,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,51200,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,64,0.004606222112973531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,65536,0.045886221859190196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,256,32,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,16384,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,65536,0.042465776205062866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,12288,0.004955555415815777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,51200,0.037053333388434515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,51200,0.03431199987729391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,16384,0.015496888094478183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,10240,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,16384,0.014151111245155334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,12288,0.013076444466908773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,8192,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,12288,0.012269333004951477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,10240,0.012030222349696688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,7168,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,256,0.0021297778520319196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,10240,0.011015999648306104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,6144,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,8192,0.009549332989586724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,7168,0.010037333601050908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,5120,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,6144,0.009497777455382878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,4096,0.004242666479614046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,6144,0.008621333373917473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,5120,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,3584,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,5120,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,4096,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,3072,0.004583111239804162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,4096,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,3584,0.007715555528799693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,2560,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,3584,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,3072,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,2048,0.004187555362780889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,3072,0.006715555571847492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,2560,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,256,512,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,2560,0.006671999891599019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,2048,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,1024,0.0028106667515304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,2048,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,1536,0.006412444429265128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,768,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,8192,0.010452444354693094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,1024,0.006016888966162999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,1536,0.005660444498062134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,512,0.0024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,1024,0.005413333161009683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,256,0.001962666710217794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,7168,0.009371555513805812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,512,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,768,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,128,0.0018035556293196147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,512,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,256,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,256,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,65536,0.0073928890956772705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,128,0.004946666873163647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,51200,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,64,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,16384,0.00535733335547977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,65536,0.041801777150895864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,12288,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,1536,0.003526222374704149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,51200,0.034252444903055825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,10240,0.0052871112194326185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,16384,0.014158222410413953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,8192,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,7168,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,12288,0.01205244412024816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,6144,0.0046942221621672315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,10240,0.011011555790901184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,8192,0.009707555174827576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,768,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,7168,0.009000889129108852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,4096,0.004201777693298128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,3584,0.00452977791428566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,6144,0.008411555654472774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,128,64,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,3072,0.0041680000722408295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,5120,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,15,128,128,0.00563733321097162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,4096,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,2048,0.004214222232500712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,3584,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,1536,0.003536000019974179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,3072,0.006555555595291986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,1024,0.002793777734041214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,128,32,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,2560,0.006304889089531369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,768,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,512,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,2048,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,256,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,1536,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,128,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,1024,0.005197333378924264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,768,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,64,0.001773333383931054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,32,0.0017964444640609955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,256,0.004620444443490771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,65536,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,128,0.004587555511130227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,64,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,5120,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,32,0.004512000001139111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,16384,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,12288,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,65536,0.04178933302561442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,51200,0.034232887956831194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,64,2560,0.004523555437723796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,16384,0.014141332772043018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,8192,0.005037333402368757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,7168,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,10240,0.011020444333553314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,8192,0.009376889301670922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,7168,0.009000000026490953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,5120,0.004184000194072723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,6144,0.008216000265545314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,64,512,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,4096,0.0039013334446483185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,5120,0.007671111159854465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,3584,0.00423911131090588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,3072,0.003945777813593547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,2560,0.003875555677546395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,51200,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,2048,0.003912000192536248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,4096,0.007000889215204451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,3584,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,1536,0.003522666792074839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,1024,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,2560,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,768,0.0024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,10240,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,2048,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,512,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,256,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,12288,0.012112889024946423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,1536,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,128,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,1024,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,64,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,32,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,768,0.00518844442235099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,512,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,15,32,6144,0.004583111239804162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,256,0.004581333448489507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,128,0.004863111095296012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,64,0.004621333132187526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,32,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,16384,0.3214328818851047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,12288,0.22613599565294054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,16384,0.16371556123097739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,15,32,3072,0.006660444455014334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,51200,0.4887973467508952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,10240,0.1882613367504544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,51200,0.9663004345364041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,8192,0.15471200148264566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,12288,0.12395644187927246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,10240,0.10313866535822551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,10240,0.08776977989408706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,7168,0.13278133339352077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,7168,0.06376889016893175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,6144,0.1144568920135498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,8192,0.11401332749260797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,8192,0.07504266500473022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,7168,0.092284447617001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,5120,0.09623910983403523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,4096,0.08054844538370769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,6144,0.08499911096360947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,6144,0.05552000138494703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,5120,0.07039378086725871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,3584,0.06821244292789035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,12288,0.12233600351545547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,4096,0.039354665411843195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,4096,0.061792890230814614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,3072,0.05932266844643486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,3584,0.05191911260286967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,2560,0.04996177885267469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,3072,0.03161511156294081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,3072,0.04792799883418613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,2048,0.0411137772930993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,2560,0.027608889672491286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,2560,0.04242577817704943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,1536,0.0316826668050554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,51200,0.43329511748419863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,2048,0.027775110469924078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,1024,0.022408889399634466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,1536,0.01921688848071628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,16384,0.16303466426001653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,1024,0.01975555552376641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,768,0.017873777283562552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,1024,0.01516888870133294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,512,0.012527111503813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,768,0.013387555877367655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,768,0.01587022178702884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,3584,0.0351857774787479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,256,0.007076444725195567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,512,0.012727999852763282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,512,0.012708444562223224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,256,0.010409778190983666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,256,0.010006222460005019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,128,0.01039644413524204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,128,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,2048,0.03463644451565213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,32,0.0041857775714662345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,64,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,65536,5120,0.04710311028692457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,1536,0.02432800001568264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,65536,32,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,128,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,65536,0.5161724620395237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,65536,1.0860320197211373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,16384,0.2793875535329183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,51200,0.8558657964070638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,12288,0.206115550465054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,16384,0.13050488630930582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,16384,0.13194578223758274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,10240,0.17273244592878553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,12288,0.09803466664420234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,51200,0.3454906675550673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,12288,0.10119199752807617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,8192,0.13883288701375326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,65536,64,0.004239999999602635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,10240,0.06918933656480578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,7168,0.12261244985792373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,8192,0.09250577953126694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,8192,0.06087377998563978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,6144,0.10438399844699436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,7168,0.07618577612770928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,7168,0.05084000031153361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,5120,0.08835022317038642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,6144,0.045808888143963285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,4096,0.07206044594446818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,5120,0.038438220818837486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,5120,0.056926223966810435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,3584,0.06318933433956571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,65536,0.44317155414157444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,4096,0.051947553952534996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,4096,0.03269955515861511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,3584,0.029030223687489826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,3072,0.05472266674041748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,3584,0.04265333215395609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,2560,0.046542220645480685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,2048,0.036954667833116316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,51200,0.39070844650268555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,2560,0.02240177823437585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,10240,0.08350844515694512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,1536,0.028928889168633357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,2048,0.01947111056910621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,1024,0.02057422200838725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,2048,0.028654221031400893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,1536,0.016135111451148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,1536,0.020229332976871066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,768,0.015135111080275642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,1024,0.013273778061072031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,6144,0.07042755683263142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,512,0.010659555594126383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,768,0.011834666960769229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,512,0.011011555790901184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,768,0.013813333378897773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,512,0.01129155523247189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,128,0.0031928889867332247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,256,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,256,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,64,0.002534222271707323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,128,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,51200,3072,0.02629866699377696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,32,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,3072,0.038252443075180054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,128,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,2560,0.035311109489864774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,64,0.008943999807039896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,32,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,65536,0.3704311052958171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,65536,0.17821155654059517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,51200,0.26851288477579754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,16384,0.08942844470342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,51200,1024,0.016315556234783597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,51200,0.131404439608256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,65536,0.15852977169884577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,12288,0.06967644559012519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,10240,0.05323288838068644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,51200,256,0.0064328888224230866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,51200,0.13190666834513345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,16384,0.04795555604828728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,12288,0.03916710946294997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,8192,0.04373777906099955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,12288,0.036517331997553505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,7168,0.038006222910351224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,10240,0.028424888849258423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,10240,0.031814220878813006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,8192,0.02628266645802392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,6144,0.03219199842876858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,5120,0.02750933335887061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,7168,0.022106667359670002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,7168,0.031033777528338965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,6144,0.019914666811625164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,6144,0.028231110837724473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,4096,0.023542222049501207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,8192,0.034634666310416326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,5120,0.017476444443066914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,3584,0.020351111888885498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,3072,0.017847110827763874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,5120,0.023694222172101338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,4096,0.015486222174432544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,3584,0.014288889037238227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,2560,0.015127110812399121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,4096,0.02117777698569828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,3072,0.01311644415060679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,3584,0.017267554998397827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,2560,0.012074666718641916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,3072,0.016480889585283067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,1536,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,2560,0.015571556157535978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,2048,0.01276533305644989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,1024,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,2048,0.013427555561065674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,1536,0.00906844437122345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,768,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,1024,0.007958222594526079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,768,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,512,0.0041724443435668945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,1536,0.011411555939250521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,768,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,256,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,512,0.006979555719428592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,128,0.002201777779393726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,256,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,512,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,256,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,128,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,32,0.0021359999146726397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,128,0.00553688903649648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,2048,0.012280888855457306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,32,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,65536,0.2522471215989855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,16384,16384,0.04262488749292162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,51200,0.1976862218644884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,65536,0.14519822597503662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,1024,0.009128889275921715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,16384,0.06531288888719347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,65536,0.1391484472486708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,12288,0.050832887490590416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,51200,0.11263999674055312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,16384,0.041511999236212835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,16384,64,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,10240,0.04311111238267687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,51200,0.10742311345206366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,8192,0.035749332772360906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,12288,0.03195644418398539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,7168,0.03140888942612542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,10240,0.02569244470861223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,10240,0.02786488996611701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,16384,64,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,8192,0.027236445082558527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,6144,0.027501333090994094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,8192,0.03706044289800856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,5120,0.024068444967269897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,7168,0.026727111803160772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,6144,0.01793599956565433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,4096,0.019567110472255282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,6144,0.025223111112912495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,5120,0.016221332881185744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,3584,0.016660445266299777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,5120,0.020415999823146395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,4096,0.014310222533014087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,3072,0.014630221658282809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,4096,0.019201777047581144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,3584,0.013265777793195514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,2560,0.01238222254647149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,3072,0.012640000217490725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,3584,0.015975111060672335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,2048,0.01036000003417333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,2560,0.011319999893506369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,3072,0.015117333994971381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,1536,0.008408000071843466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,16384,0.040692445304658674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,2560,0.014462222655614218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,2048,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,1024,0.006753777878151999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,768,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,2048,0.012277333272827996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,1536,0.00998133338159985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,1024,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,512,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,1024,0.008007110820876228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,768,0.0069368887278768755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,768,0.006992888947327931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,512,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,7168,0.01962222158908844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,128,0.002502222235004107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,512,0.006443555570311016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,256,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,64,0.0021671110557185277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,256,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,128,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,32,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,128,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,64,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,65536,0.2174302207099067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,12288,32,0.006102222121424145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,51200,0.1719528834025065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,1536,0.009654222263230218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,51200,0.10787466499540542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,65536,0.12681867016686335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,12288,256,0.0028853333658642242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,16384,0.03904533386230469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,51200,0.09593155648973252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,12288,0.04529066549407112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,16384,0.03983111182848612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,10240,0.03713511096106635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,12288,0.030299554268519085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,8192,0.031202667289310034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,10240,0.02488888965712653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,10240,0.02677244444688161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,7168,0.02777333392037286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,8192,0.021943110558721755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,8192,0.026345777842733595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,6144,0.02455377744303809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,7168,0.019313777486483257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,7168,0.023434667123688593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,65536,0.1404391129811605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,5120,0.021204445097181533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,12288,12288,0.03499911228815714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,6144,0.017490666773584154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,4096,0.01754133403301239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,16384,0.058279110325707324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,5120,0.015822221835454304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,3584,0.015141333142916361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,6144,0.020814221766259935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,5120,0.01767200065983666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,4096,0.013807111316257052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,3072,0.013459555804729462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,2560,0.01142133358452055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,4096,0.015758223003811307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,3584,0.012890666723251343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,2048,0.009746666583749983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,3072,0.011591110792424945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,3584,0.014409777190950183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,1536,0.008021333151393467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,2560,0.01038755559258991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,1024,0.006552888701359431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,2048,0.009240888886981541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,2560,0.012297777665985955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,1536,0.00816888858874639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,2048,0.010949333508809408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,1536,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,512,0.0037955554823080697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,1024,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,768,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,256,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,768,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,512,0.006911110960774952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,128,0.002472888885272874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,512,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,256,0.0063635553750726916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,64,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,256,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,32,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,128,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,128,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,64,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,65536,0.1722853316201104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,32,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,51200,0.13008533583747015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,65536,0.0880897773636712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,3072,0.013081777426931592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,65536,0.07220444414350721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,16384,0.044234666559431285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,10240,12288,0.03056000007523431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,12288,0.035014222065607704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,10240,1024,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,51200,0.056391113334231906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,10240,0.026710222164789837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,16384,0.0270453327231937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,12288,0.02019733356104957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,12288,0.021583999196688335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,10240,0.016601777738995023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,7168,0.020793777373101976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,10240,0.018942221999168396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,8192,0.015154666370815702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,8192,0.020782222350438435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,6144,0.018202667435010273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,7168,0.013411555853154926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,7168,0.017854221993022494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,5120,0.015457777513398064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,6144,0.01239911135700014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,6144,0.01700444519519806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,5120,0.01165511127975252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,5120,0.014871110518773397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,10240,768,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,51200,0.06936711072921753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,3584,0.01146755533085929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,4096,0.013478221992651621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,16384,0.02437422176202138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,3584,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,8192,0.023522666758961145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,3584,0.011510222322411008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,2560,0.008960000342792934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,3072,0.008896888958083259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,3072,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,2048,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,2560,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,1536,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,2560,0.010370666782061258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,2048,0.008012444608741337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,1024,0.004854222138722737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,2048,0.009184000392754873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,1536,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,768,0.003555555724435382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,1536,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,1024,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,1024,0.006607999818192587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,4096,0.013582222163677216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,768,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,512,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,768,0.006217777729034424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,256,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,4096,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,128,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,512,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,64,0.0018737777653667664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,3072,0.010430222584141625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,256,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,32,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,8192,128,0.005969777703285217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,256,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,128,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,65536,0.14691466755337185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,64,0.005418666534953647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,8192,32,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,51200,0.116230222913954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,16384,0.04048711061477661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,65536,0.08360622326533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,65536,0.06661155488755968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,12288,0.0317520002524058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,51200,0.05372800098525154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,51200,0.06281510988871257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,16384,0.02198311189810435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,10240,0.02674666709370083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,16384,0.025062221619817946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,8192,0.02314666741424137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,12288,0.017887110511461895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,12288,0.019893333315849304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,7168,0.019484443797005545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,10240,0.015607110328144498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,6144,0.01702222228050232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,8192,0.01408088869518704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,10240,0.01756533318095737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,5120,0.014554666148291694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,8192,0.02072177827358246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,4096,0.012449777788586087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,6144,0.012093333734406365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,6144,0.015640000502268474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,3584,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,5120,0.01370666672786077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,4096,0.009583111438486312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,3072,0.009887110855844285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,4096,0.012778666284349231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,3584,0.009010666774378883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,2560,0.008646222452322641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,3584,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,3072,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,2048,0.007253333098358578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,3072,0.011047999891969891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,2560,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,1536,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,2560,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,2048,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,1024,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,2048,0.008634666601816813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,1536,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,768,0.0036640001667870414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,1536,0.006665777828958299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,8192,512,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,1024,0.006246222390068903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,512,0.0028888889484935333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,1024,0.006330666856633292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,256,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,512,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,7168,0.01699733402993944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,768,0.005982222242487802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,128,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,5120,0.011049778097205691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,512,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,64,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,256,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,256,0.005179555465777715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,128,0.005780444376998478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,128,0.004977777600288391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,65536,0.12216356065538193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,64,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,7168,32,0.005294222384691238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,51200,0.09607822365230984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,65536,0.07444088988833956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,65536,0.06344622373580933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,16384,0.03421333432197571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,51200,0.056871996985541455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,12288,0.02706844442420536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,16384,0.02084355552991231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,16384,0.023694222172101338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,10240,0.023297778434223596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,7168,0.013033777475357056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,12288,0.01719466679626041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,8192,0.01981511049800449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,10240,0.015240889456537036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,10240,0.016003555721706815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,7168,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,7168,0.01771111124091678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,8192,0.01367555558681488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,8192,0.020098666350046795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,7168,0.01257777793539895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,7168,0.016194666425387066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,5120,0.013766222529941134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,6144,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,6144,0.015188443991872998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,4096,0.01074399964676963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,5120,0.010433778166770935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,5120,0.013076444466908773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,51200,0.051595555411444775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,3584,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,4096,0.008754666480753157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,4096,0.012088889049159156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,3072,0.009277332988050249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,3584,0.008341333104504479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,7168,768,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,2560,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,3584,0.010738666686746808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,3072,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,2048,0.006899555524190267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,12288,0.018237334158685472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,1536,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,2048,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,2560,0.009326222042242685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,1024,0.004002666721741359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,2048,0.007676444119877285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,6144,0.015799111790127225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,1536,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,1536,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,768,0.0033155555526415506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,512,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,1024,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,256,0.0022106667359670005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,1024,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,768,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,128,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,512,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,768,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,64,0.0018719999740521114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,256,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,6144,32,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,128,0.005680000202523337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,256,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,128,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,65536,0.106312890847524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,64,0.005212444398138259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,32,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,51200,0.08406577507654826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,16384,0.030839999516805012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,65536,0.06710755825042725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,65536,0.061305774582756885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,3072,0.01020088874631458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,51200,0.05123466584417555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,16384,0.02019466625319587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,16384,0.02127200033929613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,10240,0.02100177771515316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,12288,0.016580444243219163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,8192,0.018538667096032035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,12288,0.016873778568373788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,10240,0.014811555544535318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,7168,0.016083555089102853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,10240,0.014826666977670459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,8192,0.01312888910373052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,6144,0.014301333162519666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,8192,0.01828444500764211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,6144,512,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,7168,0.012436444560686747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,5120,0.01240799989965227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,7168,0.014821334017647637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,4096,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,6144,0.014223110344674854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,5120,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,5120,0.012440000143316058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,3584,0.009561777942710454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,4096,0.008542222281297049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,4096,0.011425777441925473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,3072,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,6144,2560,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,3584,0.008278222547637092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,51200,0.04961955547332764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,2560,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,3584,0.010267555713653564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,3072,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,2048,0.006562666760550604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,2560,0.007368000017272101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,3072,0.008672888908121321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,2048,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,2048,0.007672000262472365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,1024,0.0035386664999855887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,1536,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,1536,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,768,0.003232888877391815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,1024,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,6144,0.011077333655622272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,1024,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,768,0.006095110956165526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,768,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,256,0.0022622222701708474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,512,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,512,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,128,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,12288,0.024795555406146582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,256,0.00564533347884814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,64,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,256,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,32,0.0018088887963030075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,5120,128,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,128,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,65536,0.08543288707733154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,2560,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,64,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,5120,32,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,51200,0.06633422109815809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,16384,0.025144888295067683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,65536,0.060025778081681996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,65536,0.05938844548331367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,51200,0.046943111552132495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,12288,0.020280000236299302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,16384,0.019482667247454327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,512,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,10240,0.01757333344883389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,16384,0.01887822151184082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,12288,0.015999111864301894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,8192,0.015450666348139444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,12288,0.015522667103343539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,10240,0.014358222484588623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,7168,0.01290844463639789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,10240,0.013750221994188098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,6144,0.01145066652033064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,8192,0.012714666624863943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,5120,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,8192,0.017138666576809354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,6144,0.010216000179449717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,4096,0.009204444785912832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,5120,0.008904889225959778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,6144,0.013441777891582914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,3584,0.007788444558779399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,5120,1536,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,5120,0.011868444581826528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,3072,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,4096,0.008191111187140146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,4096,0.010990222295125326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,3584,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,2560,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,3072,0.007532444265153673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,3584,0.00908622228437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,2048,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,3072,0.008276444342401292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,51200,0.04760444495413038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,1536,0.00388355553150177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,2048,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,1024,0.0031822222388452957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,2048,0.007024889190991719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,1536,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,1024,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,768,0.002871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,1024,0.006028444402747684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,512,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,7168,0.011696889168686338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,7168,0.014154665999942355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,768,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,256,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,512,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,512,0.005296000176005893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,256,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,128,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,32,0.0018968888454967076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,2560,0.007272889216740926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,2560,0.00793688909875022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,64,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,1536,0.006339555399285422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,4096,32,0.004952888935804367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,51200,0.05921600262324015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,65536,0.05768711037105984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,65536,0.0581279993057251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,16384,0.02221155497762892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,768,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,51200,0.04487288991610209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,51200,0.04692000150680542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,12288,0.017880888448821176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,16384,0.01921688848071628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,4096,128,0.0019484443797005548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,10240,0.015476443701320224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,16384,0.018556444181336295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,8192,0.013399110900031196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,12288,0.015978667471143935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,10240,0.014335110783576965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,7168,0.012048000262843238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,10240,0.013330666555298699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,8192,0.01641244524055057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,6144,0.010964444114102257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,7168,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,65536,0.07619288894865248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,5120,0.009649777577983009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,6144,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,6144,0.012934221989578672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,4096,0.008743111458089616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,5120,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,5120,0.011373333632946014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,3584,0.007585778004593319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,4096,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,4096,0.009296889106432596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,3072,0.006894222150246303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,3584,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,3584,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,2560,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,3072,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,12288,0.014968888627158271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,3072,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,2048,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,8192,0.012135999898115793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,2560,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,1536,0.003965333518054751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,2048,0.0069857777820693115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,1024,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,2048,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,1536,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,7168,0.013711999687883588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,768,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,1536,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,1024,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,512,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,768,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,1024,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,256,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,128,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,768,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,256,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,512,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,64,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3584,32,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,256,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,128,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,65536,0.061701330873701304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,64,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,32,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,51200,0.04909155435032315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,4096,128,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,65536,0.05736266904407077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,16384,0.019262222780121695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,65536,0.05526399943563673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3584,2560,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,51200,0.04576622115241157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,51200,0.04342133469051785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,10240,0.01514755520555708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,16384,0.01789955629242791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,12288,0.015676443775494892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,12288,0.014498665928840637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,8192,0.013412444127930535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,10240,0.013697778185208639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,7168,0.012057777908113269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,10240,0.01311822235584259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,6144,0.010960000256697336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,8192,0.015229332778188916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,7168,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3584,128,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,5120,0.009008888569143083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,7168,0.012732444538010491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,4096,0.008263111114501953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,6144,0.012069332930776807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,5120,0.010002666877375709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,3584,0.007253333098358578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,4096,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,4096,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,3072,0.00657155571712388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,12288,0.015808888607554965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,3584,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,2560,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,16384,0.018908444378111098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,3584,0.007671111159854465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,3072,0.007532444265153673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,3072,0.007290666302045186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,2048,0.00461777796347936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,2560,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,1536,0.0037919998996787598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,2560,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,2048,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,1024,0.0031991110493739447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,2048,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,8192,0.011870221959220039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,768,0.0028693332440323303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,1536,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,1024,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,1536,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,768,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,6144,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,1024,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,256,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,768,0.00535733335547977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,5120,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,128,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,512,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,64,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,256,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,512,0.005393777870469623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,32,0.0018613334331247543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,256,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,65536,0.056631108125050865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,128,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,51200,0.04390933447413974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,32,0.004946666873163647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,65536,0.056127998563978404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,16384,0.019819556011093985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,65536,0.05314844515588549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,51200,0.04522577921549479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,51200,0.041872001356548734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,12288,0.01646311084429423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,16384,0.018777777751286823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,10240,0.014129777749379476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,16384,0.01721600029203627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,12288,0.015397333436542086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,8192,0.012895999683274163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,12288,0.014279999666743808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,10240,0.013080889152155982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,3072,512,0.002464888824356927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,7168,0.011232000258233813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,10240,0.01274222218328052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,6144,0.010086221827401055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,8192,0.011044444309340583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,7168,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,8192,0.01291822228166792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,5120,0.008856000171767341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,7168,0.011935110721323224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,6144,0.009669333696365356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,4096,0.007880888879299164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,6144,0.010370666782061258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,3584,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,5120,0.008632000121805403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,4096,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,3072,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,3584,0.008007110820876228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,3584,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,2560,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,3072,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,3072,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,2048,0.004287111262480418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,2560,0.007436444362004598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,1536,0.0037493333220481873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,2560,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,2048,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,1024,0.0032026666320032547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,2048,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,1536,0.006344888773229387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,768,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,1536,0.006024888820118374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,1024,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,512,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,3072,128,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,1024,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,768,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,5120,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,256,0.0022195554855797025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,768,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,128,0.0018915554715527429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,4096,0.008278222547637092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,256,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,512,0.005304000029961268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,128,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,32,0.0018231111268202465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,256,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,65536,0.04797866609361437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,64,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,32,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,51200,0.035544888840781316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,65536,0.05544711152712504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,16384,0.016179554992251925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,65536,0.05050488975312975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,51200,0.04459111226929558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,16384,0.018536888890796237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,51200,0.04041422075695462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,12288,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,16384,0.016727111405796476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,10240,0.011890666352377998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,12288,0.014717333846622042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,12288,0.013786666923099093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,8192,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,10240,0.012725333372751871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2560,512,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2560,64,0.0018595556418100994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,8192,0.011026666396194033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,8192,0.012152888708644442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,6144,0.00904088881280687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,7168,0.01035377797153261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,7168,0.010021333065297868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,5120,0.008223999705579545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,6144,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,6144,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,4096,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,5120,0.00903733323017756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,5120,0.008070222205585903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,3584,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,4096,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,3584,0.008005333443482717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,3072,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,10240,0.012529777983824411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,3072,64,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,7168,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,3072,0.007687110867765214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2560,128,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,3072,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,2048,0.0042239998777707415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,2560,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,1536,0.0038737778862317405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,2560,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,2048,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,1024,0.003208000005947219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,1536,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,2048,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,768,0.0028320000403457214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,1536,0.006005333529578314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,1024,0.006060444646411472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,512,0.002510222295920054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,1024,0.005624889085690181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,768,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,256,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,768,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,4096,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,512,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,128,0.0018906665758954154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,64,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,512,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,32,0.0020186667227082783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,3584,0.007328000333574083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,128,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,256,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,65536,0.03477155499988132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,2048,2560,0.004874666531880696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,51200,0.027622222900390625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,64,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,16384,0.012721777790122561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,65536,0.055009779002931386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,65536,0.04891022377543979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,51200,0.044440888696246676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,12288,0.011237333218256632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,16384,0.01685066686736213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,10240,0.009926222264766693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,16384,0.01640888883007897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,12288,0.013174222575293647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,8192,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,12288,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,10240,0.011895111037625207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,10240,0.011756444142924415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,7168,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,8192,0.010992888775136737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,6144,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,8192,0.010312889185216691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,7168,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,5120,0.007084444165229797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,7168,0.009394666386975182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,6144,0.009383999639087254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,2048,256,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,4096,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,6144,0.008795555267069075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,3584,0.00620977787507905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,5120,0.008991999758614434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,5120,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,128,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,4096,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,3072,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,2560,0.004858666823969947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,4096,0.007624000310897827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,3584,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,2048,0.004319111092223061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,3072,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,3584,0.007376000285148621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,1536,0.003499555504984326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,2560,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,3072,0.007002666592597961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,51200,0.03871733281347487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,2560,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,1536,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,2048,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,768,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,1536,0.005965333431959152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,1024,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,1024,0.005496888938877318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,512,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,768,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,256,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,768,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,512,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,128,0.0019137778629859288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,512,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,256,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,64,0.0017991111510329777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,128,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,32,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,2048,32,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,64,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,65536,0.027871112028757732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,32,0.004855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,51200,0.021856889128684998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,2048,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,65536,0.054400000307295054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,16384,0.010622222390439777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,65536,0.046851555506388344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,51200,0.04384266667895847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,12288,0.008585777547624376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,51200,0.03769066598680284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,16384,0.015853333804342482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,10240,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,16384,0.01534933348496755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,12288,0.01311466677321328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,8192,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,12288,0.012431110772821637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,10240,0.011741333537631564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1536,256,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,7168,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,10240,0.011086222198274402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,6144,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,8192,0.010515555739402771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,7168,0.00998399986161126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,8192,0.01001422190003925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,5120,0.006514666808976068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,7168,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,6144,0.009373333719041612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,4096,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,6144,0.008815111385451423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,3584,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,4096,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,5120,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1536,1024,0.002985777747299936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,4096,0.007401777638329401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,3584,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,3584,0.007322666545708974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,3072,0.007255111303594377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,3072,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,2048,0.0069288888739215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,2560,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,2560,0.006555555595291986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,1536,0.004946666873163647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,2048,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,1024,0.0038853333228164246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,2048,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,1536,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,768,0.0035502223504914176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,1536,0.006064888917737537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,1024,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1536,128,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,512,0.0028799999919202593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,768,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,1024,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,256,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,768,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,512,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,128,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,256,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,512,0.0052879999081293745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,3072,0.005428444594144821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,128,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,32,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,65536,0.022031999296612207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,128,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,2560,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,51200,0.01759466694460975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,64,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,32,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,65536,0.05396355523003472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,16384,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,65536,0.04618488748868307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,51200,0.04344355397754245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,12288,0.00794400026400884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,51200,0.037455999188952975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,16384,0.015840000576443143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,10240,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,16384,0.015116444892353482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,12288,0.013078221844302284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,8192,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,10240,0.01202933324707879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,12288,0.012259555359681448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,7168,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,10240,0.011062222222487131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,8192,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,8192,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,6144,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,7168,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,5120,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,7168,0.009373333719041612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,1024,64,0.0019066666977273093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,6144,0.009539555344316695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,4096,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,1024,256,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,6144,0.008766222331258986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,3584,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,5120,0.00868800034125646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,3072,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,5120,0.008218666745556725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,3584,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,4096,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,2560,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,3584,0.00683111117945777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,3072,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,2048,0.0041751112374994485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,3072,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,2560,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,1536,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,2560,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,2048,0.006709333509206772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,1024,0.003583999971548716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,2048,0.005991999887757831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,768,0.0031946665710873077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,1536,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,1024,0.005375999957323074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,512,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,768,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,1024,5120,0.008605333666006723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,256,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,512,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,512,0.004946666873163647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,128,0.0019511110666725372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,4096,0.008261333737108443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,256,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,64,0.0017991111510329777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,128,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,128,0.004583111239804162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,768,32,0.001869333287080129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,65536,0.015767110718621146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,64,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,768,32,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,1536,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,51200,0.013159111142158508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,65536,0.05048711101214091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,16384,0.0075866662793689305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,51200,0.03862488932079739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,768,0.00563733321097162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,12288,0.006918222126033571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,16384,0.015504888362354703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,10240,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,51200,0.03696444299485948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,16384,0.014483556151390076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,12288,0.013163555827405719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,10240,0.012063999970753988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,12288,0.012127111355463663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,7168,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,10240,0.011019555230935415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,6144,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,768,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,8192,0.009718221922715506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,7168,0.010006222460005019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,5120,0.006413333531883027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,6144,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,4096,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,5120,0.008613333106040955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,6144,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,3584,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,5120,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,4096,0.007724444071451823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,3072,0.0052844443255000645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,4096,0.007112888826264276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,3584,0.007667555577225155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,65536,0.045824001232783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,2560,0.005200888961553574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,3584,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,2048,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,3072,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,2560,0.006990222467316522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,1536,0.003559111307064692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,2560,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,8192,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,1024,0.0029102222373088202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,1536,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,1536,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,8192,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,768,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,1024,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,512,0.002463111033042272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,1024,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,768,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,7168,0.00922400007645289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,256,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,768,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,512,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,512,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,128,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,256,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,64,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,256,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,512,32,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,65536,0.01072355525361167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,128,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,64,0.004862222406599256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,3072,0.006741333338949416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,51200,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,65536,0.046612444851133555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,512,2048,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,16384,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,65536,0.04246311055289375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,2048,0.006180444525347815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,12288,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,51200,0.03734577695528666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,16384,0.015842666228612263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,10240,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,51200,0.034288889831966825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,8192,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,12288,0.01277511070171992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,10240,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,12288,0.012070222033394707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,10240,0.011057777537239922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,8192,0.010519111322032081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,6144,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,8192,0.00965244405799442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,7168,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,7168,0.009081777599122789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,5120,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,6144,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,4096,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,6144,0.00849777791235182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,5120,0.008660444782839881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,3584,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,5120,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,4096,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,3072,0.004591111093759537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,4096,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,512,32,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,3584,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,2560,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,3584,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,2048,0.004570666700601578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,3072,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,3072,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,2560,0.007204444044166141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,1536,0.003519999898142285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,2048,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,2560,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,16384,0.014298665854665967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,1024,0.002891555635465516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,768,0.002529777793420686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,1536,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,7168,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,512,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,1024,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,1536,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,256,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,768,0.0058319999112023254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,1024,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,128,0.0019351111518012153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,512,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,768,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,64,0.0018275556051068837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,256,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,256,32,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,256,128,0.005635555419656966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,256,0.005045333256324132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,65536,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,51200,0.007125332951545715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,128,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,64,0.004621333132187526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,16384,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,65536,0.04588355620702108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,32,0.004549333204825719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,12288,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,65536,0.04251022140185038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,51200,0.03386133246951633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,16384,0.015807110402319167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,10240,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,16384,0.014440000057220459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,12288,0.013086222112178802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,12288,0.011984888878133563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,10240,0.011730666789743634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,7168,0.005034666508436203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,10240,0.011046222514576383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,8192,0.010421333213647207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,8192,0.00979022185007731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,7168,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,7168,0.00907466643386417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,5120,0.004586666822433472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,6144,0.009317333499590555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,512,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,5120,0.008661333057615492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,4096,0.0042035554846127825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,51200,0.037308444579442344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,4096,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,4096,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,3584,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,8192,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,3072,0.004574222283230888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,3584,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,3584,0.006715555571847492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,256,2048,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,2560,0.00452533322903845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,6144,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,3072,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,3072,0.006817777951558431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,2048,0.0041911109454101985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,1536,0.003535110917356279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,2560,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,2048,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,1024,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,6144,0.008842666943868002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,2048,0.006061333335108227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,1536,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,768,0.00254133323000537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,1536,0.0058266665372583605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,512,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,1024,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,5120,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,768,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,1024,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,256,0.001975999938117133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,128,0.0018088887963030075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,768,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,256,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,512,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,64,0.001790222194459703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,128,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,128,32,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,256,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,65536,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,51200,0.00693866651919153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,128,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,16384,0.005295111073387994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,64,0.004589333302444882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,12288,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,32,0.0046311111913787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,10240,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,65536,0.04173955652448866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,51200,0.03366044494840834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,16384,0.01443911095460256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,7168,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,6144,0.00462755560874939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,10240,0.01071911139620675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,128,2560,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,5120,0.004568888909286923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,8192,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,4096,0.0038844446341196695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,7168,0.009318222602208456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,3584,0.00423555572827657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,3072,0.004192000048028098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,6144,0.008341333104504479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,2560,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,4096,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,14,128,512,0.00574933323595259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,2048,0.0041866666740841335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,3584,0.007222221957312689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,1536,0.0035111109415690103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,3072,0.006568888823191325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,1024,0.0028080000645584534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,768,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,2560,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,512,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,2048,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,1536,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,1024,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,128,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,512,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,8192,0.005205333232879639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,64,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,32,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,128,0.0047004446387290955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,12288,0.011791110866599612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,65536,0.00832799987660514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,51200,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,64,0.004707555390066571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,32,0.004578666554556953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,65536,0.04177422324816386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,5120,0.00793599999613232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,12288,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,10240,0.005087999833954705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,16384,0.014144000079896716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,8192,0.004951999833186467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,12288,0.012134222520722283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,7168,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,10240,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,6144,0.004610666798220741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,64,256,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,8192,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,5120,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,768,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,4096,0.004293333325121137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,3584,0.004552000098758274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,7168,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,3072,0.004230222354332606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,6144,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,64,256,0.004944889081848992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,2560,0.0041271112859249115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,5120,0.007864889171388414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,2048,0.003879111260175705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,4096,0.007002666592597961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,1536,0.0031840000301599503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,3584,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,1024,0.0028693332440323303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,768,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,3072,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,16384,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,2560,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,2048,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,256,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,1536,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,128,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,64,0.0018079999006456798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,1024,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,32,0.0018071112119489247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,51200,0.033583998680114746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,512,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,256,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,128,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,64,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,32,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,16384,0.32396621174282497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,16384,0.1629813379711575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,51200,0.4888906478881836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,16384,0.1367057826783922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,51200,0.9655022091335721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,14,32,512,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,10240,0.18773867024315727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,12288,0.12191022766960992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,14,32,768,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,8192,0.1534551117155287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,7168,0.13265866703457302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,8192,0.0718417763710022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,12288,0.10387111372417873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,8192,0.11261599593692356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,10240,0.10302488671408759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,6144,0.11432711283365886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,51200,0.4052346547444661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,5120,0.09633511304855347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,7168,0.06373955806096394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,12288,0.226295100318061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,6144,0.0832408865292867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,4096,0.08035822047127618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,3584,0.0683457785182529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,5120,0.07873244418038262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,5120,0.04714222086800469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,4096,0.039697776238123574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,3072,0.059283554553985596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,3584,0.03490488727887472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,4096,0.05817511346605089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,3584,0.04959288901752896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,2560,0.050027555889553495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,3072,0.0310204459561242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,2048,0.04080444574356079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,2560,0.03188177943229675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,3072,0.04888355400827196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,10240,0.088319996992747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,2560,0.037652442852656044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,1024,0.022799111074871484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,2048,0.023063111636373732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,1536,0.019038221902317472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,1536,0.02548355526394314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,768,0.01777866648303138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,1024,0.014985778265529208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,6144,0.05493333273463779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,7168,0.09616000122494167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,512,0.012545777691735161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,1024,0.01940622263484531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,256,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,768,0.013291555146376291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,128,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,512,0.012847999731699625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,768,0.016128000285890367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,256,0.010553778045707278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,64,0.0041955556306574075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,32,0.004210666649871402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,65536,128,0.010637332995732626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,128,0.008262222011884054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,64,0.010027555955780877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,32,0.010024888647927178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,65536,1536,0.03173244330618117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,2048,0.035972442891862653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,65536,0.5146062109205458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,65536,1.0893324746025932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,51200,0.39112623532613117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,51200,0.8472640249464246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,16384,0.27696000205145943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,256,0.009769777456919352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,12288,0.20753155814276802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,12288,0.09792888826794094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,12288,0.08437422249052261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,65536,0.44669686423407656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,16384,0.13073955641852483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,51200,0.33237600326538086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,10240,0.17296177811092803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,8192,0.1383262210422092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,10240,0.08319466643863253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,7168,0.1224222183227539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,10240,0.0710560017161899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,65536,512,0.0124195557501581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,8192,0.0573857757780287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,8192,0.08879999981986152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,6144,0.10444888803693984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,7168,0.07630755503972371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,5120,0.08924711412853664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,7168,0.05122133427196079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,4096,0.07209244701597425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,6144,0.06348088714811537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,5120,0.03927110963397556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,5120,0.06342222293217976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,4096,0.053773333628972374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,4096,0.03256355391608344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,16384,0.11190577348073323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,3584,0.029153777493370905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,3584,0.03979644510481093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,2560,0.04589777853753832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,3072,0.02641244398223029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,3072,0.03917600048912896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,2048,0.03713599840799967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,2560,0.02274755636850993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,1536,0.028954668177498713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,2560,0.0313351121213701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,2048,0.023412444525294836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,1024,0.020413332515292697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,2048,0.029778665966457788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,1536,0.016660445266299777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,1536,0.02238133384121789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,768,0.014977777997652689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,1024,0.013449778159459433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,512,0.0105591110057301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,1024,0.016576889488432143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,6144,0.04457955559094747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,256,0.006464889066086874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,768,0.01200177768866221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,3584,0.06286133660210504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,128,0.0035084444615576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,256,0.009364444348547194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,512,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,3072,0.054970665110482105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,64,0.0025457777082920074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,256,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,51200,32,0.002817777709828483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,128,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,128,0.007270221908887227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,64,0.00903911143541336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,32,0.008962666822804345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,65536,0.3700586689843072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,65536,0.17412889003753662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,51200,0.2686435646480984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,16384,0.08925866418414646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,51200,0.13129511144426134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,65536,0.15375644630855984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,12288,0.06956089205212064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,16384,0.047705776161617704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,51200,768,0.014133334159851074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,10240,0.05319111214743721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,12288,0.033349331882264875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,12288,0.03642844491534763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,8192,0.04351022177272373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,7168,0.03786044319470724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,10240,0.03155466583040025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,10240,0.029010666741265193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,8192,0.024617777930365667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,8192,0.040023999081717596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,7168,0.022272000710169475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,5120,0.0281644430425432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,6144,0.01977333260907067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,6144,0.02757599949836731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,4096,0.023766222927305434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,5120,0.017616000440385606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,5120,0.02351466649108463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,3584,0.020523554748959012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,4096,0.015491555134455362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,4096,0.019896888070636325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,3072,0.017553778158293832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,3584,0.014306666122542487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,51200,0.11647110515170628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,3584,0.01721600029203627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,2560,0.014902222487661572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,3072,0.013162666724787818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,3072,0.01756355497572157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,2048,0.01238666640387641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,2560,0.012145777543385824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,1536,0.00965244405799442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,2560,0.014085332552591959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,2048,0.010886222124099731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,1024,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,7168,0.03673955467012193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,1536,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,2048,0.013425778183672162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,768,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,1536,0.011387555963463254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,1024,0.007730666961934831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,512,0.0041777777175108595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,768,0.007326222128338284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,768,0.007768889268239339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,256,0.002546666603949335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,512,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,16384,0.042360001140170626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,128,0.0022053333620230355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,51200,512,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,256,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,64,0.0021280000607172647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,256,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,32,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,128,0.006678222368160884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,128,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,16384,6144,0.03288622366057502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,64,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,32,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,65536,0.2515164481268989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,51200,0.19776266151004365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,65536,0.14544799592759874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,16384,0.06530221965577868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,16384,1024,0.009223110973834991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,65536,0.1361102263132731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,51200,0.11268088552686904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,12288,0.0507244434621599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,51200,0.10944800244437324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,16384,0.04132711225085788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,16384,512,0.007330666813585494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,16384,0.03772977656788296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,12288,0.029631111356947158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,8192,0.035940445131725736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,12288,0.03197777933544583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,7168,0.03133599956830343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,10240,0.02570399973127577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,6144,0.027552889453040227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,10240,0.027894222074084814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,8192,0.03046400017208523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,7168,0.01977777812216017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,5120,0.02392533255947961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,7168,0.02738044493728214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,6144,0.018009778526094224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,4096,0.01923822197649214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,6144,0.023970666858885024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,5120,0.016128889388508268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,3584,0.0166284441947937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,5120,0.022061333060264587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,4096,0.014388443695174323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,3072,0.01475822263293796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,3584,0.01351999988158544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,2560,0.012287110918098025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,3584,0.015965332587560017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,4096,0.01943644384543101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,2048,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,3072,0.015941333439615037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,1536,0.008562666674455007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,2560,0.01124533348613315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,10240,0.0429368880059984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,2560,0.01315822203954061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,1024,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,2048,0.012412444584899478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,768,0.00554044461912579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,1536,0.010419555836253695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,1024,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,8192,0.021872889664438035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,1024,0.00833777752187517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,512,0.0037644443412621817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,256,0.0028311111446883944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,768,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,512,0.00702311098575592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,128,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,256,0.0063288890653186375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,512,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,64,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,12288,32,0.0022035555707083806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,128,0.0063973334100511335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,128,0.0057911111248864075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,3072,0.012416889270146688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,64,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,65536,0.21688266595204672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,12288,32,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,51200,0.17190399434831408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,2048,0.01111822244193819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,65536,0.14173955387539333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,1536,0.008635555704434713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,16384,0.057913780212402344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,12288,0.045244445403416954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,65536,0.12778755029042563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,51200,0.10692977905273438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,51200,0.0988053348329332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,16384,0.03624177641338772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,10240,0.03695999913745456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,8192,0.03074399961365594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,12288,768,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,12288,0.028667555914984807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,7168,0.027623999449941847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,12288,0.03048711021741231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,10240,0.026700443691677515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,6144,0.024717332588301763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,8192,0.026675555441114638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,7168,0.019142222073343065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,5120,0.021095110310448542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,6144,0.017502221796247695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,6144,0.020199111766285367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,5120,0.015838222371207345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,4096,0.01773955590195126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,5120,0.01777777738041348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,3584,0.01514933341079288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,4096,0.013887999786270989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,4096,0.016113777955373127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,3072,0.013409777647919126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,3584,0.012814222110642327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,3584,0.01384622189733717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,2560,0.01144088887506061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,3072,0.011896000140243106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,16384,0.03939022289382087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,2048,0.009766221874290043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,3072,0.013755555782053204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,10240,0.024667556087176006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,2560,0.01110222190618515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,8192,0.021222222182485793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,1536,0.008057778080304464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,2560,0.011615999870830111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,1024,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,2048,0.009318222602208456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,768,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,7168,0.025069332785076563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,2048,0.01108444482088089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,1536,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,512,0.003793777690993415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,1536,0.009229333036475712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,1024,0.007377777662542131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,768,0.006872889068391588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,256,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,128,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,512,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,768,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,64,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,512,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,256,0.006521777974234686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,10240,32,0.0021768889079491296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,128,0.0063635553750726916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,256,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,128,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,65536,0.17188533147176108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,32,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,51200,0.1290257771809896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,16384,0.04426133301522997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,65536,0.08957155545552571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,65536,0.07119288709428576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,12288,0.03477955526775784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,51200,0.05698755714628431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,51200,0.06954844130410089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,16384,0.02295733326011234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,10240,0.026752000053723652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,16384,0.027790221903059218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,8192,0.024421332610978022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,12288,0.018650665879249573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,12288,0.0214675549003813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,10240,0.01629866659641266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,10240,1024,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,10240,0.0188711103465822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,8192,0.014759999182489185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,8192,0.020191111498408847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,7168,0.013443555268976422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,5120,0.01554666625128852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,6144,0.012635555532243518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,7168,0.01848266687658098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,4096,0.01368444412946701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,5120,0.011630222201347351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,6144,0.016016888949606154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,3584,0.011728000309732227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,4096,0.010438222024175854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,5120,0.014396443963050842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,10240,64,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,3072,0.010327111515733931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,2560,0.008956444760163626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,4096,0.013271999855836233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,3584,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,3072,0.008816888762844933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,3584,0.01107377807299296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,2560,0.008385777473449707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,3072,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,2560,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,2048,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,1024,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,2048,0.009017777939637503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,1536,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,768,0.003662222375472387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,1536,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,1024,0.00656088896923595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,7168,0.020733333296246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,512,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,1024,0.006302222195598815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,6144,0.018016889691352844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,768,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,256,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,512,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,768,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,128,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,256,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,512,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,64,0.0019093333846992916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,32,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,256,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,128,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,65536,0.14892621835072836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,64,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,8192,32,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,2048,0.007600888609886169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,51200,0.11582844787173802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,8192,1536,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,16384,0.0404151115152571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,65536,0.08272355794906616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,12288,0.03162577748298645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,51200,0.0537413325574663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,51200,0.06289600001441108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,16384,0.021586666504542034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,16384,0.025456888808144465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,12288,0.017768000562985737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,8192,0.023026666707462732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,12288,0.019876443677478366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,10240,0.015822221835454304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,7168,0.019525332583321463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,10240,0.017314665847354464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,8192,0.014022222823566861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,6144,0.016788444585270353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,8192,0.019646222392718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,8192,128,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,7168,0.012979555461141797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,5120,0.014719999498791166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,7168,0.01751733322938283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,6144,0.012079111403889127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,4096,0.01242844429281023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,6144,0.015815999772813585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,5120,0.011071110765139261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,3584,0.011018666956159802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,5120,0.01346577786737018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,4096,0.00983911090426975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,3072,0.009923555784755284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,3584,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,2560,0.008581333690219456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,3584,0.011139555937714048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,3072,0.008277333445019191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,2048,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,65536,0.06637599733140734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,3072,0.010823999842007955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,2560,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,1536,0.006208000083764394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,10240,0.026722666290071275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,2560,0.009633777870072259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,2048,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,768,0.003551111039188173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,2048,0.00830488900343577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,1024,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,512,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,768,0.006045333213276333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,1024,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,768,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,256,0.002500444443689452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,512,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,128,0.002120888895458645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,512,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,256,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,64,0.0018639999131361644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,256,0.0052888890107472735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,128,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,32,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,128,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,65536,0.12134222189585368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,64,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,4096,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,32,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,51200,0.09581955273946126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,65536,0.07417244381374784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,16384,0.034168889125188194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,7168,1024,0.004459555364317364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,65536,0.0636977752049764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,51200,0.056903998057047524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,12288,0.027071999178992376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,7168,1536,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,51200,0.051579554875691734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,10240,0.023318222827381555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,16384,0.02069155540731218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,7168,1536,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,8192,0.020021332634819877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,12288,0.017200888858901132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,16384,0.023783110909991797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,7168,0.017665778597195942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,12288,0.018170666363504197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,6144,0.015815999772813585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,8192,0.013761777844693927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,10240,0.01611288885275523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,5120,0.013819555441538492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,7168,0.012640888492266336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,8192,0.018414222531848483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,4096,0.010928000013033548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,7168,0.016135111451148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,6144,0.011714666253990598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,5120,0.010718222293588849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,3584,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,6144,0.014390221900410123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,3072,0.009280000295903947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,5120,0.013091555900043912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,4096,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,2560,0.0079644446571668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,4096,0.012074666718641916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,3584,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,2048,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,3072,0.007797333101431529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,3584,0.010743111371994019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,1536,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,2560,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,2048,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,2560,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,1024,0.004278222305907143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,1536,0.006567111031876669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,2048,0.008114666574531132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,768,0.0032675556010670135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,1536,0.007159111400445302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,512,0.0028719999310043124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,768,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,1024,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,256,0.0022320000247822865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,768,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,10240,0.015307555596033732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,512,0.005884444548024072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,512,0.005646222167544895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,256,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,64,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,256,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,128,0.005659555395444234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,32,0.0018906665758954154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,128,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,65536,0.10667200220955743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,64,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,32,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,51200,0.08392533328798081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,6144,3072,0.010425777898894416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,65536,0.06116533279418945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,16384,0.030846221579445735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,12288,0.024726221958796184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,51200,0.05132888754208883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,51200,0.04971111151907179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,16384,0.020218667056825426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,10240,0.02091644373204973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,6144,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,12288,0.016632889707883198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,12288,0.01657777859105004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,10240,0.014924445086055331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,7168,0.01610933409796821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,10240,0.014820444915029736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,8192,0.013275555438465543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,6144,0.014176888598336114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,8192,0.017298666967286002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,7168,0.012187555432319641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,5120,0.012390221986505719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,7168,0.015124445160230001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,6144,0.011361777782440186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,4096,0.0107004443804423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,6144,0.01349688900841607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,5120,0.010142222046852112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,3584,0.009508444203270806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,5120,0.012999110751681857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,4096,0.008974221845467886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,3072,0.008599110775523717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,65536,0.06749155786302355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,4096,0.011622221933470832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,3584,0.007928888830873702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,2560,0.0075706665714581805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,3072,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,2048,0.006543111056089401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,2560,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,16384,0.02128711177243127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,2560,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,1536,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,8192,0.01833511061138577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,2048,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,2048,0.007663999994595845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,1536,0.006766222417354584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,768,0.0034142221427626083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,1024,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,768,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,512,0.0028435554769304064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,768,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,512,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,256,0.0022986665782001284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,512,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,3584,0.010007999837398529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,256,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,128,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,3072,0.009644444617960189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,256,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,64,0.0018231111268202465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,32,0.002011555557449659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,128,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,5120,1024,0.0037582222786214617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,128,0.004972444640265571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,1536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,65536,0.08561066786448161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,64,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,5120,32,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,5120,1024,0.006483555667930179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,16384,0.025026667449209426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,65536,0.05922400289111667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,65536,0.0614524417453342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,51200,0.047544889979892306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,51200,0.046915554338031344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,16384,0.01953866746690538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,16384,0.019158222609096102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,10240,0.01755555636352963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,12288,0.016142222616407607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,12288,0.015374221735530429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,10240,0.014464888307783337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,10240,0.013804444836245643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,7168,0.012878222597969903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,8192,0.01275644451379776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,8192,0.016306666864289176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,6144,0.011366222467687396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,7168,0.011622221933470832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,7168,0.015113777584499784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,6144,128,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,5120,0.009970666633711921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,6144,0.01034400032626258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,6144,0.013090666797426013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,4096,0.008978666530715095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,51200,0.0661804477373759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,5120,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,3584,0.007754666937722101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,4096,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,5120,0.012792888614866467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,3072,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,12288,0.020264888803164165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,4096,0.010713777608341642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,3584,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,2560,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,2048,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,3584,0.008721777962313758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,3072,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,1536,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,2560,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,3072,0.009329777624871995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,8192,0.01535555554760827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,2048,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,1024,0.0031742221779293488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,2560,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,768,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,2048,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,1536,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,512,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,768,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,1536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,256,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,1024,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,768,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,128,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,512,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,64,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,256,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,512,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,4096,32,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,128,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,256,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,128,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,51200,0.05909866756863064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,64,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,4096,32,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,65536,0.05818488862779406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,65536,0.058000889089372426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,51200,0.04683377676539951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,51200,0.045145776536729604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,12288,0.017902221944597032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,10240,0.015461333923869662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,16384,0.01926666663752662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,16384,0.018213333355055917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,12288,0.015955555770132277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,8192,0.013354666531085968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,12288,0.014814222852389017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,10240,0.01443911095460256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,7168,0.011993777420785693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,10240,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,8192,0.012118221984969245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,6144,0.011001778145631155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,4096,1024,0.006179555422729916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,8192,0.015497777197096082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,5120,0.00963288876745436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,7168,0.011161777708265515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,4096,0.008658666577604082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,7168,0.01387288918097814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,6144,0.01016444464524587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,3584,0.007431999676757389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,5120,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,4096,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,3072,0.006834666762087081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,3584,0.007932444413503012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,65536,0.07592978080113728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,3584,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,2560,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,3072,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,3072,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,2560,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,2048,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,2560,0.006918222126033571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,1536,0.0038702223036024305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,2048,0.0069875551594628235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,2048,0.007032000356250339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,1024,0.0031395554542541504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,1536,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,1536,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,768,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,1024,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,1024,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,6144,0.012178666889667511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,512,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,768,0.005980444451173146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,5120,0.011382222175598145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,256,0.002193777718477779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,768,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,128,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,512,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,256,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,512,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,64,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,16384,0.022219555245505437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3584,128,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,256,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,128,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,65536,0.061836441357930504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,51200,0.049060444037119545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,64,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,32,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,16384,0.01923644377125634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,65536,0.057168887721167676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,65536,0.05552355448404948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,12288,0.015498666299713982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,16384,0.01904800037542979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,16384,0.01751733322938283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,10240,0.015234667393896313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,12288,0.015606222881211175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,12288,0.014442667365074158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,8192,0.013400000002649097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,10240,0.013745778136783175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,10240,0.013088888592190213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3584,4096,0.009653333160612319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,7168,0.012048888537618848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,8192,0.011775111158688864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,8192,0.013412444127930535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,6144,0.01094133324093289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3584,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,7168,0.010761777559916178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,5120,0.009026666482289633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,7168,0.012854222622182635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,6144,0.009538667069541084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,4096,0.00831733312871721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,5120,0.008967111508051554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,6144,0.011030221978823343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,3584,0.007249777515729268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,5120,0.00981244444847107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,3072,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,4096,0.008353778057628209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,2560,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,3584,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,4096,0.008234666453467475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,3072,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,51200,0.04336977667278714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,3072,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,1536,0.0038933331767717996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,2560,0.007035555938879649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,2048,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,1024,0.0031911111954185697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,2048,0.00657155571712388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,1536,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,768,0.0028604444944196274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,1024,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,1024,0.005758222192525864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,768,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,512,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,768,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,256,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,512,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,512,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,128,0.0020782223178280727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,256,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,256,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,64,0.001818666648533609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,128,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,2048,0.004576888763242298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,128,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,3584,0.007277333074145847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,2560,0.007041778001520369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,65536,0.05671377976735433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,51200,0.04330044322543674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,32,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,65536,0.05608533488379585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,65536,0.05300177799330818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,16384,0.01976444489426083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,1536,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,51200,0.04532088836034139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,51200,0.04190399911668566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,12288,0.016453334026866488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,16384,0.01889600025282966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,10240,0.014111111561457316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,16384,0.01702844434314304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,12288,0.015057777365048727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,8192,0.012803555362754397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,12288,0.014109333356221518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,10240,0.013420444395807056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,10240,0.012864888542228274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,7168,0.011318222516112857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,8192,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,6144,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,8192,0.013229333692126803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,3072,51200,0.04607999987072415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,7168,0.01036711119943195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,3072,32,0.0018177777528762817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,7168,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,6144,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,4096,0.007914666500356462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,6144,0.01033688916100396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,5120,0.008720888859695857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,3584,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,5120,0.008952889177534316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,4096,0.008316444853941599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,3072,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,3584,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,2560,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,3584,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,3072,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,2048,0.0042453333735466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,3072,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,2560,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,1536,0.0038728887836138406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,2560,0.006909333169460297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,2048,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,1024,0.0031973332580592898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,2048,0.006305777778228124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,1536,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,768,0.0028239999794297745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,1536,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,1024,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,512,0.0024604443460702896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,1024,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,768,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,5120,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,512,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,128,0.0019048889064126545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,512,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,256,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,64,0.001793777777089013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,3072,64,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,32,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,256,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2560,128,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,128,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,4096,0.007479111353556315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,51200,0.035531553957197405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,64,0.004942222187916438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,32,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,65536,0.05540000067816841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,16384,0.016173332929611206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,65536,0.05074400040838453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,51200,0.044622222582499184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,12288,0.013217777841620974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,51200,0.04020266731580099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,16384,0.016749334004190233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,12288,0.01444355481200748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,12288,0.013806222213639153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,8192,0.01092888911565145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2560,256,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,10240,0.012400888734393649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2560,768,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,7168,0.009647111097971598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,8192,0.011065777805116443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,6144,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,8192,0.012487110992272695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,7168,0.01035288886891471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,7168,0.01017955525053872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,65536,0.049245334333843656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,6144,0.009335111412737105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,4096,0.007247111035717859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,6144,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,5120,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,5120,0.00834222220712238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,3584,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,4096,0.00831200016869439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,3072,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,4096,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,3584,0.007806222471925948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,16384,0.018221333622932434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,2560,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,10240,0.011795555551846823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,3584,0.007246221933099959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,3072,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,2048,0.004311999926964442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,1536,0.0038408889538711975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,2560,0.007017778025733099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,2048,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,2560,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,1024,0.0031866667171319327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,10240,0.012423999607563019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,2048,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,768,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,1536,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,1024,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,512,0.002471111093958219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,768,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,1024,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,5120,0.00814133303032981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,768,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,512,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,128,0.0018951110541820526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,256,0.005535111245181825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,256,0.004944889081848992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,128,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,64,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,128,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,65536,0.03485511077774896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,64,0.005043555465009478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,3072,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,32,0.004959110998445087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,51200,0.02753066685464647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,65536,0.04862133330769009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,65536,0.05487199955516391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,16384,0.012666666673289405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,2048,1536,0.00637511122557852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,12288,0.011379555695586734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,51200,0.044320887989468045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,51200,0.03902666767438253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,16384,0.017145777742067974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,10240,0.010017777482668558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,12288,0.013435555828942193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,2048,256,0.0021626665774318907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,8192,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,12288,0.013408888545301227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,10240,0.012145777543385824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,7168,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,10240,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,8192,0.010750222537252637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,6144,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,8192,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,5120,0.007234666910436418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,7168,0.009368889033794403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,6144,0.009302222066455418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,4096,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,6144,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,5120,0.009030222064918941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,3584,0.005979555762476391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,4096,0.008345777789751688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,4096,0.007277333074145847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,3072,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,3584,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,3584,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,2560,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,3072,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,3072,0.006923555499977536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,2048,0.0041946665280395085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,16384,0.01648355523745219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,2560,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,1536,0.003559111307064692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,2048,0.006607999818192587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,2560,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,1024,0.002979555477698644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,2048,512,0.005111111121045219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,2048,0.006025777922736273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,768,0.002809777855873108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,7168,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,1536,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,1024,0.006105777704053455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,768,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,1024,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,256,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,768,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,128,0.0019048889064126545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,512,0.005292444593376583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,256,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,5120,0.008273777862389883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,64,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,32,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,128,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,128,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,65536,0.02755022214518653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,64,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,32,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,65536,0.05444000164667765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,65536,0.04690666662322151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,16384,0.010325333310498131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,51200,0.04383111000061035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,51200,0.037639111280441284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,1536,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,16384,0.015837333268589444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1536,512,0.002461333241727617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,10240,0.0075902218619982404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,12288,0.013183111117945777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,12288,0.01236977759334776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1536,512,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,8192,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,10240,0.012053333222866058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,10240,0.011117333339320289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1536,256,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,7168,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,8192,0.010687111152542962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,8192,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,6144,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,7168,0.00998133338159985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,5120,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,51200,0.021903999977641638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,7168,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,6144,0.009671111073758867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,4096,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,6144,0.008893333375453949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,3584,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,5120,0.008651555412345463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,5120,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,4096,0.008303999900817871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,3072,0.005537777725193236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,4096,0.007408888803588019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,3584,0.007624000310897827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,12288,0.008443555898136562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,2560,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,2048,0.005687110953860813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,3584,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,3072,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,1536,0.004859555512666702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,16384,0.015579554769727917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,2560,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,3072,0.0069439998931354946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,1024,0.0038853333228164246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,2048,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,2560,0.006447111152940326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,768,0.0032133333798911837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,1536,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,512,0.0028640000770489373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,1024,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,1536,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,256,0.002430222307642301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,1024,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,512,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,768,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,128,0.0021359999146726397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,256,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,512,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,64,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,256,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,1024,32,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,128,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,65536,0.021819555097156104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,128,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,51200,0.01751733322938283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,64,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,32,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,65536,0.05414577656322055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,65536,0.04599644409285652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,12288,0.007894222107198503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,51200,0.04311911265055338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,16384,0.01585244470172458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,51200,0.03715022073851691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,10240,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,16384,0.014581332604090372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,12288,0.013101333545313941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,8192,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,12288,0.012158222496509552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,10240,0.012057777908113269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,7168,0.006890666567616993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,1024,2048,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,10240,0.011010666688283285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,8192,0.010363555616802638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,6144,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,5120,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,1024,768,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,7168,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,4096,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,7168,0.009307555854320526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,5120,0.00869599978129069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,6144,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,5120,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,4096,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,3072,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,4096,0.007032000356250339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,3584,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,2560,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,3584,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,3072,0.007328000333574083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,2048,0.004184000194072723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,16384,0.009295110901196798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,3072,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,2560,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,2048,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,2560,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,1024,0.0038693332009845306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,2048,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,1536,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,768,0.0029253332565228143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,1024,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,1024,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,512,0.0025617778301239014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,8192,0.009683555199040307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,768,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,6144,0.009331555830107795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,256,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,512,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,3584,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,128,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,512,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,256,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,64,0.0017964444640609955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,256,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,32,0.0018702221827374564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,768,128,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,65536,0.01584533353646596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,128,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,64,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,768,1536,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,51200,0.013237333132161034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,65536,0.0483804444471995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,16384,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,65536,0.04535999894142151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,51200,0.03860800133811103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,12288,0.006867555694447623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,1536,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,51200,0.03701511025428772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,16384,0.01585244470172458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,10240,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,16384,0.014639111028777229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,8192,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,12288,0.01313333296113544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,12288,0.012085333466529846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,768,0.0052871112194326185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,7168,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,8192,0.010448888772063784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,10240,0.011083555718262991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,6144,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,8192,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,5120,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,6144,0.009047110875447592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,7168,0.009283555878533257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,4096,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,6144,0.008420444197124904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,5120,0.00870311094654931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,5120,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,3584,0.005623111294375525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,4096,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,3072,0.005295111073387994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,4096,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,768,32,0.004860444615284602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,3584,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,2560,0.004976888911591636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,3584,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,3072,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,2048,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,2560,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,3072,0.006746666712893381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,1536,0.0035342222286595237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,2560,0.00639111093348927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,2048,0.006895111252864202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,1024,0.0028533333291610083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,1536,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,2048,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,768,0.002805333377586471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,10240,0.012030222349696688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,1024,0.005995555470387141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,1536,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,512,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,256,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,768,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,1024,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,128,0.0019537777536445195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,512,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,768,0.005052444421582752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,512,32,0.0018231111268202465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,512,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,128,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,65536,0.010689777632554373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,256,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,128,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,51200,0.009312888814343346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,64,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,65536,0.04628533456060621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,512,32,0.004861333303981357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,65536,0.042351110113991626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,51200,0.03731111023161147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,12288,0.005192888693677055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,51200,0.03490133417977227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,16384,0.015799111790127225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,10240,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,16384,0.014130666851997375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,8192,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,10240,0.01202755586968528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,12288,0.01219466659757826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,7168,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,10240,0.01108000013563368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,8192,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,6144,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,8192,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,7168,0.009920000202125972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,7168,0.009284444153308868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,5120,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,6144,0.009073778159088558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,6144,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,4096,0.005212444398138259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,5120,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,3584,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,5120,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,4096,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,3072,0.004701333327425851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,16384,0.005336888962321811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,4096,0.0073724447025193115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,2560,0.004971555537647671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,3584,0.007679999702506595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,3072,0.007223111059930589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,2048,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,3072,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,2560,0.006996444529957241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,12288,0.013048888908492194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,1536,0.003531555748648114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,2560,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,1024,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,2048,0.0069022224181228215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,2048,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,1536,0.006593777901596493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,768,0.0026506665680143568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,1536,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,1024,0.006311111152172089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,512,0.002458666761716207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,768,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,1024,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,256,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,768,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,512,7168,0.009992889232105678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,128,0.0018595556418100994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,512,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,256,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,64,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,256,32,0.0017911110901170308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,256,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,65536,0.0079644446571668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,128,0.004537777768241035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,64,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,51200,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,32,0.004865777575307422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,65536,0.04588355620702108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,16384,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,256,3584,0.006907555378145642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,65536,0.04243377844492594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,12288,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,51200,0.03733866744571262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,10240,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,16384,0.015837333268589444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,51200,0.03388799892531501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,12288,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,8192,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,10240,0.011749332977665795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,7168,0.004945777770545748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,8192,0.010405333505736457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,10240,0.011056000159846412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,6144,0.004860444615284602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,8192,0.010021333065297868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,7168,0.009662222531106737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,512,0.005542222410440445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,5120,0.0046160001721647055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,7168,0.009093333449628618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,6144,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,4096,0.004476444588767158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,6144,0.00888000014755461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,5120,0.008627555436558193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,3584,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,256,128,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,3072,0.004598222259018156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,4096,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,5120,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,2560,0.004515555583768421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,3584,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,3072,0.007323555648326874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,3584,0.007103111180994246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,2048,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,2560,0.007247111035717859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,3072,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,1536,0.0034968890249729156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,2560,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,2048,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,1536,0.006565333240562015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,2048,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,16384,0.014256000518798828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,1536,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,12288,0.011986666255527072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,1024,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,512,0.002510222295920054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,1024,0.0052879999081293745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,768,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,768,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,512,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,128,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,256,0.0052826665341854095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,256,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,128,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,64,0.001810666587617662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,128,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,4096,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,32,0.0017831110292010838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,65536,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,64,0.004547555413511065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,51200,0.006639110959238476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,128,32,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,1024,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,16384,0.005326222214433882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,65536,0.042400889926486544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,12288,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,10240,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,768,0.0025395556456512874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,8192,0.004982222285535601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,51200,0.033611555894215904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,7168,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,12288,0.012079999678664737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,6144,0.0047182221379545005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,10240,0.011064888702498542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,13,128,512,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,8192,0.009649777577983009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,5120,0.004597333156400257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,4096,0.0041831110914548235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,3584,0.004373333520359463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,7168,0.008829333715968663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,3072,0.0041911109454101985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,6144,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,2560,0.004558222161398994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,5120,0.007631111476156447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,4096,0.007328888608349695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,2048,0.004252444538805219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,1536,0.0033315554675128725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,3584,0.006994666324721442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,1024,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,3072,0.006588444527652528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,768,0.002518222149875429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,2560,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,512,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,1536,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,256,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,1024,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,128,0.0018622221218215094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,768,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,64,0.0017911110901170308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,512,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,64,32,0.001782222133543756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,256,0.004567111117972268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,16384,0.014094221923086377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,128,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,128,256,0.002104000084929996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,51200,0.007270221908887227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,64,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,32,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,12288,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,65536,0.041783110962973706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,51200,0.03358577688535055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,10240,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,16384,0.014159111513031853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,12288,0.012103999654452005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,7168,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,10240,0.010749333434634738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,6144,0.004576888763242298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,8192,0.009641778137948778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,64,2048,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,5120,0.004519999855094486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,4096,0.004278222305907143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,7168,0.009024000002278222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,6144,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,5120,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,3072,0.004232888834344016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,4096,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,2560,0.004032888760169347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,2048,0.003863999827040566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,3584,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,1536,0.0034924443397257063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,65536,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,1024,0.0027466666781240036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,16384,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,3072,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,2560,0.00666844430896971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,768,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,512,0.002199999988079071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,256,0.0019022222194406723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,1536,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,128,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,1024,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,64,0.0018151110659042995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,768,0.005151999907361137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,512,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,32,0.001828444500764211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,256,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,128,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,64,0.004555555681387584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,32,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,3584,0.004530666602982415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,16384,0.32317066192626953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,51200,0.48816442489624023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,16384,0.16230132844713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,13,32,2048,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,51200,0.9643013212415906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,13,32,8192,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,12288,0.2258817884657118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,16384,0.1591493288675944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,12288,0.12216177251603867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,10240,0.18777333365546334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,12288,0.10361511177486843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,51200,0.4629191292656793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,8192,0.15384356180826822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,10240,0.1282542281680637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,10240,0.1117511060502794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,8192,0.11460622151692708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,8192,0.07134666707780626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,16384,0.1746844450632731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,7168,0.13258311483595106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,12288,0.13192889425489637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,6144,0.11451821857028538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,7168,0.10239289204279582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,7168,0.063646223809984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,10240,0.11373155646853977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,5120,0.0961644450823466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,6144,0.0855368905597263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,8192,0.09556711382336086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,6144,0.055396444267696805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,4096,0.08027911186218262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,5120,0.07775911357667711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,5120,0.04736622174580892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,7168,0.0835440026389228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,3584,0.06811022096210055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,4096,0.06549421946207683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,6144,0.07389866643481784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,5120,0.06360355350706312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,3584,0.052523556682798594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,4096,0.039388444688585066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,3584,0.03497599893146091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,4096,0.053652445475260414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,2560,0.049997333023283214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,3072,0.031368888086742826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,3584,0.0490880012512207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,3072,0.04938222302330864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,2048,0.04090311129887899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,2560,0.03956800036960178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,3072,0.04398577743106418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,2560,0.03319999906751845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,1536,0.031621333625581526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,2560,0.038663112454944186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,2048,0.03488533364401923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,2048,0.023041778140597876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,1024,0.022612444228596155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,1536,0.01910044418440925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,2048,0.03407377666897244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,1536,0.026313778426912095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,768,0.018009778526094224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,1536,0.029675556553734675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,1024,0.019976889093716938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,512,0.01272266689274046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,1024,0.02306755549377865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,768,0.013497778111033969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,768,0.016176889340082806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,256,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,3072,0.059177776177724205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,768,0.02001066671477424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,512,0.012846222354306115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,512,0.012695111334323883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,128,0.005286222116814719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,512,0.016215110818545025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,256,0.010598222414652506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,64,0.004192000048028098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,256,0.009963555468453301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,65536,32,0.0041724443435668945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,256,0.013554666605260638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,128,0.010479999913109673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,128,0.008100444244013892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,128,0.010093332992659675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,32,0.009840000006887648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,65536,64,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,65536,1024,0.0151591118839052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,65536,0.517772462632921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,51200,0.39030665821499294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,65536,1.0900755988226998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,51200,0.8481653531392416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,16384,0.27640533447265625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,65536,51200,0.5038977728949653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,65536,0.42616889211866593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,16384,0.13062132729424372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,12288,0.20757244692908394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,16384,0.12595644262101915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,51200,0.3759360048505995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,12288,0.0980079968770345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,10240,0.17347378200954863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,12288,0.08218489090601604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,10240,0.10245866907967462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,8192,0.14016800456576878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,10240,0.09092177947362264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,16384,0.14164533880021837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,8192,0.09003111388948228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,7168,0.12216888533698188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,8192,0.05777066946029663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,12288,0.1069599986076355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,10240,0.09675466352038914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,7168,0.08025778002209134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,6144,0.1046346690919664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,8192,0.07772799995210436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,7168,0.05118311113781399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,5120,0.08904800150129531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,6144,0.07049777772691515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,6144,0.04458222124311659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,7168,0.0693004462454054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,4096,0.071578668223487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,5120,0.06353866391711764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,6144,0.0634746683968438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,4096,0.0329039990901947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,5120,0.053300443622801036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,3584,0.06287466817431979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,4096,0.044616887966791786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,3584,0.029106666644414265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,3584,0.04278666774431864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,3584,0.04210844304826525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,3072,0.05509066581726074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,51200,0.3992302152845595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,3072,0.04159644577238295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,3072,0.025927111506462097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,5120,0.03838222225507101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,2560,0.0458133320013682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,3072,0.037598222494125366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,2048,0.03711288836267259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,2560,0.022623111804326374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,4096,0.05107466710938347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,2048,0.029304888513353135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,1536,0.028757333755493164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,2048,0.019318222999572754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,2560,0.034113778008355036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,2048,0.02998488810327318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,1024,0.02069155540731218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,65536,0.5193520122104222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,1536,0.016466667254765827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,768,0.015127110812399121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,1536,0.022664889693260193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,1536,0.025311110748185053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,1024,0.016716443830066256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,512,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,768,0.011985777980751462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,768,0.017557332913080852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,1024,0.01994755533006456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,512,0.011113777756690979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,256,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,512,0.011028444601429833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,768,0.01401600076092614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,128,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,256,0.009343999955389235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,512,0.015184889237085978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,64,0.0028106667515304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,256,0.008373333348168267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,51200,32,0.0028328889360030494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,2560,0.033850666549470686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,128,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,128,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,128,0.009322666459613377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,51200,256,0.012379555238617791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,64,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,51200,32,0.008912888665994009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,65536,0.17577511734432646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,51200,1024,0.013439111411571503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,51200,0.1309022241168552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,16384,0.08949155277676052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,65536,0.15625066227383083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,51200,0.1348088847266303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,16384,0.048103110657797925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,16384,0.042585776911841504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,12288,0.06952177815967135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,16384,0.056001775794559054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,12288,0.036567999256981745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,12288,0.03339733348952399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,10240,0.053301334381103516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,12288,0.04362044400639004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,8192,0.043362667163213096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,10240,0.04007911019855075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,10240,0.03607644306288825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,51200,0.27116889423794216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,10240,0.03903466794225905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,51200,0.13766577508714464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,7168,0.03794044587347243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,8192,0.02438755498992072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,8192,0.035177777210871376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,6144,0.03303822212749057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,65536,0.18667022387186685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,8192,0.03456977672047085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,7168,0.0314044455687205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,7168,0.022456000248591106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,5120,0.028202666176689997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,6144,0.019908444748984445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,6144,0.02844711144765218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,4096,0.022873777482244704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,5120,0.01762488815519545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,6144,0.028734223710166082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,5120,0.025210666987631056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,3584,0.020548444655206468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,5120,0.026022222306993272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,4096,0.01556799974706438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,4096,0.0226248883538776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,3584,0.014474666780895658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,3072,0.01757866640885671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,3584,0.018068444397714403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,3584,0.02102311121092902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,3072,0.013136000268989138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,2560,0.014864888456132678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,65536,0.36812710762023926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,3072,0.017165333032608032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,2048,0.012434666355450949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,3072,0.01977777812216017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,2560,0.014753777119848462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,2560,0.012352888782819113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,2560,0.01808888879087236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,2048,0.013600000076823764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,2048,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,2048,0.0169831116994222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,7168,0.031178666485680476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,1024,0.0070062221752272705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,1536,0.009402666654851701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,1536,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,1536,0.015280889140235053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,768,0.006059555543793573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,1024,0.009312888814343346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,1024,0.007809778054555257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,4096,0.021526222427686054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,512,0.004201777693298128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,1024,0.013893332746293811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,768,0.007391999993059371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,768,0.007875555919276344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,256,0.0025191110455327565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,768,0.011944889194435544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,512,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,128,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,512,0.01070844464831882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,256,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,256,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,64,0.0021297778520319196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,128,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,128,0.00553955551650789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,32,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,128,0.008716444174448649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,64,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,16384,32,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,65536,0.25157067510816783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,65536,0.14757688840230307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,51200,0.19727022118038604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,65536,0.13456622759501138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,16384,512,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,51200,0.11212089326646592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,16384,0.06531466378106011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,16384,256,0.010924444430404238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,16384,0.04181866513358223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,51200,0.11230311128828262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,12288,0.05098488926887512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,16384,1536,0.009719111025333405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,16384,0.045129779312345714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,16384,0.042769776450263135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,10240,0.042817778057522245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,12288,0.031910220781962075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,8192,0.03570222192340427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,12288,0.03613244493802389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,10240,0.02663466665479872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,10240,0.03687289026048448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,7168,0.031183110343085393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,51200,0.10713777939478557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,10240,0.03288266725010342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,8192,0.022015111313925848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,6144,0.027307555079460144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,8192,0.030457778109444514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,8192,0.029539555311203003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,7168,0.020004444652133517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,7168,0.028195556667115953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,5120,0.023899555206298828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,7168,0.026708443959554035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,6144,0.017993777990341187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,6144,0.02568799919552273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,4096,0.019116444720162284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,6144,0.02401777770784166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,5120,0.01603822244538201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,5120,0.02260622216595544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,3584,0.01667555504375034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,5120,0.02164177762137519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,4096,0.0144533332851198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,4096,0.0192577772670322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,4096,0.019308444526460435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,3584,0.013353777428468069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,3584,0.016520889268981088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,12288,0.029793779055277508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,3584,0.018742221924993727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,65536,0.13840444882710776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,3072,0.012392889294359418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,3072,0.01611111064751943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,2048,0.010668444136778513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,3072,0.017487999465730455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,2560,0.011371555427710215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,2560,0.013408000270525614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,1536,0.00851822230550978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,2048,0.01275022245115704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,2048,0.009771555662155151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,2048,0.015509333875444202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,1024,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,1536,0.00978488889005449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,1536,0.01035199976629681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,1536,0.014168888330459595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,768,0.005200888961553574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,1024,0.007669332954618666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,1024,0.008010666403505538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,1024,0.01240711079703437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,512,0.004048888882001241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,3072,0.01479288935661316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,768,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,768,0.011595555477672152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,256,0.0029315555261241067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,2560,0.01236977759334776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,512,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,512,0.010395555860466426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,128,0.002528888897763358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,256,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,256,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,256,0.010684444672531553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,2560,0.01645955608950721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,64,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,128,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,12288,32,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,128,0.005299555758635203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,12288,128,0.009222221871217092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,65536,0.21579110622406006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,64,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,65536,0.13796444733937582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,51200,0.17170843813154432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,768,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,65536,0.12627822822994655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,12288,512,0.007094221810499827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,51200,0.10729422171910603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,16384,0.0579013360871209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,51200,0.10221155484517415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,12288,0.044922666417227856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,16384,0.035798221826553345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,16384,0.03912711143493652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,16384,0.04052266809675429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,10240,0.036975112226274275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,12288,0.027412444353103638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,65536,0.1206924385494656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,12288,0.03053599927160475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,8192,0.030906667311986286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,12288,0.03310755557484097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,10240,0.026875555515289307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,10240,0.030894223186704848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,7168,0.027619555592536926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,10240,0.02973244587580363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,8192,0.02647022240691715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,6144,0.024485333098305598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,8192,0.02719555629624261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,7168,0.02276355524857839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,7168,0.024331554770469666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,6144,0.02163200080394745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,6144,0.017314665847354464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,5120,0.020917332834667627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,6144,0.02242044442229801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,4096,0.017312000195185345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,5120,0.015632000234391954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,5120,0.01853155593077342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,5120,0.020222221811612446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,3584,0.014861333701345654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,4096,0.016390222642156813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,4096,0.018684445155991446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,12288,32,0.006029333505365584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,3584,0.014196443888876172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,3072,0.013414222333166333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,3584,0.012805333567990197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,8192,0.02095111045572493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,3584,0.018223111828168232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,2560,0.011376000112957425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,3072,0.011764444410800934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,7168,0.021568889419237774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,3072,0.01349866638580958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,2048,0.00963644435008367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,3072,0.0173333336909612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,2560,0.010415111151006486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,2560,0.012091555529170565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,1536,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,2560,0.016473778420024447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,2048,0.009526222116417354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,1024,0.006240889016124938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,2048,0.010875555376211802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,1536,0.00831822223133511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,2048,0.015100444356600443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,1536,0.0096142225795322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,768,0.005306666509972678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,1536,0.01402311192618476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,1024,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,512,0.004187555362780889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,4096,0.013747555514176687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,1024,0.011398221883508893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,768,0.006888888776302338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,768,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,256,0.002805333377586471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,768,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,512,0.006725333217117522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,51200,0.09407555394702488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,512,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,128,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,256,0.006230222268237009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,256,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,64,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,256,0.01000711073478063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,10240,32,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,128,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,128,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,64,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,65536,0.17275022135840523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,32,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,51200,0.1290008889304267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,65536,0.08897599909040664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,10240,1024,0.007553777760929531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,65536,0.07027377684911092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,16384,0.04412800073623657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,51200,0.06868177652359009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,51200,0.056700441572401256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,16384,0.027479999595218237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,10240,512,0.01038933296998342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,12288,0.034640000926123724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,10240,128,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,65536,0.11113333702087402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,16384,0.023427555958429973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,10240,0.0266977780395084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,16384,0.03762755460209317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,51200,0.08976978063583374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,12288,0.021703999903466966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,8192,0.024306666519906785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,10240,0.018295110927687753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,12288,0.03190044562021891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,10240,0.02369600037733714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,7168,0.020748444729381137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,10240,0.028540445698632136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,8192,0.014635556273990206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,8192,0.020562666985723708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,6144,0.01812266641192966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,8192,0.024920889072948035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,7168,0.01349511080318027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,7168,0.018039110634062026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,5120,0.015660444895426433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,7168,0.023285332653257582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,6144,0.012637333737479316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,5120,0.011674666570292579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,6144,0.02160177793767717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,4096,0.012984000146389008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,5120,0.014659555421935188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,3584,0.011509333219793109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,5120,0.019905777441130746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,4096,0.010592888626787396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,4096,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,3072,0.01033333357837465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,3584,0.009711110757456886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,4096,0.018075555562973022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,3584,0.017856000198258292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,2560,0.008718222379684448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,3072,0.008781332936551835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,3072,0.011066666907734342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,3072,0.01686844395266639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,2560,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,2560,0.010239111052619087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,2560,0.01575644479857551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,1536,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,2048,0.008203555312421586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,2048,0.009525333013799455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,2048,0.01475288967291514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,6144,0.017630222770902846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,1024,0.004851555658711327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,1536,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,1536,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,768,0.00388711111413108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,1536,0.013733333183659447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,1024,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,1024,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,512,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,1024,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,768,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,768,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,256,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,3584,0.011815999945004782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,512,0.006219555520349079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,12288,0.018328888548745047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,512,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,512,0.010454222559928894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,256,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,2048,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,64,0.001999999913904402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,128,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,32,0.0019697778754764134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,128,0.008214222060309516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,64,0.005434666656785541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,8192,32,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,65536,0.14921422799428305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,65536,0.08284711175494723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,65536,0.06589777602089776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,51200,0.11548444959852432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,768,0.011459555890825061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,8192,128,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,256,0.0058746664888328975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,51200,0.06292533212237887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,16384,0.0402817792362637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,8192,256,0.010680889089902243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,51200,0.05372888843218485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,8192,128,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,12288,0.03169688913557265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,65536,0.11036800013648139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,16384,0.021877333521842957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,16384,0.025380444195535447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,10240,0.026671111583709717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,12288,0.017696888910399545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,51200,0.08931910991668701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,16384,0.03808000021510654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,12288,0.019850666324297588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,8192,0.023072888453801472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,12288,0.03169511093033685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,10240,0.01567911108334859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,7168,0.019575999842749704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,10240,0.0225591113169988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,8192,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,8192,0.019863999552196927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,6144,0.01692800058258904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,7168,0.012868444124857584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,7168,0.018560889694425795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,5120,0.014668444792429606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,7168,0.023041778140597876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,6144,0.015827554795477126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,6144,0.012070222033394707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,6144,0.02126844392882453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,4096,0.012453333371215396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,5120,0.011183111204041375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,5120,0.01512888901763492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,3584,0.011012444065676795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,5120,0.019623999794324238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,4096,0.009758222434255812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,4096,0.012416889270146688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,3072,0.01019022199842665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,4096,0.01779733267095354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,3584,0.008948444492287105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,3584,0.011653333074516721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,2560,0.008641777767075432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,3584,0.01753777762254079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,3072,0.00832622249921163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,3072,0.011218667030334473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,2048,0.007248000138335758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,3072,0.016211556063758004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,2560,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,2560,0.010006222460005019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,10240,0.028415110376146104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,2048,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,2560,0.015472888946533203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,2048,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,1024,0.004404444247484207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,8192,0.02517066730393304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,2048,0.01497688889503479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,768,0.0036799998746977914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,1536,0.006741333338949416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,1536,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,1024,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,1536,0.013074666261672974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,512,0.0028684443483750024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,1024,0.006552000012662675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,768,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,768,0.006225777582989798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,512,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,768,0.010397333237859938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,512,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,128,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,512,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,256,0.005639111002286275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,256,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,64,0.0018888889915413326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,256,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,32,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,128,0.004952888935804367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,128,0.00830488900343577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,64,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,1536,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,7168,32,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,65536,0.07589155435562134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,65536,0.06356355879041883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,7168,1024,0.0110871113008923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,7168,256,0.0023493334237072202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,51200,0.056632002194722496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,16384,0.03409689002566867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,65536,0.11096711291207208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,51200,0.05167466733190748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,12288,0.0269777774810791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,16384,0.023419555690553453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,16384,0.02095111045572493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,7168,128,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,16384,0.037336889240476824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,10240,0.022896000080638464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,12288,0.01720177796151903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,12288,0.018388445178667705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,65536,0.12148355113135444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,8192,0.019883554842736986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,10240,0.02251200046804216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,12288,0.03120888935195075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,10240,0.027315555347336665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,51200,0.09545777903662787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,7168,0.01757777730623881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,8192,0.01367644468943278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,8192,0.018557333283954196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,8192,0.02437688906987508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,6144,0.015461333923869662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,7168,0.012535110943847232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,7168,0.016502221425374348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,7168,0.022314666046036616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,5120,0.013800000150998434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,6144,0.011702222128709158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,6144,0.014914666612943014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,4096,0.010685332947307162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,6144,0.02124622298611535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,5120,0.010354667074150508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,5120,0.013619555367363824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,51200,0.08922577566570705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,3584,0.009322666459613377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,5120,0.019288889235920377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,4096,0.009200000100665623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,4096,0.01204533295498954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,3072,0.009293333523803288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,4096,0.01792711185084449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,3584,0.008623110751310984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,2560,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,3584,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,3072,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,10240,0.015299555328157214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,3072,0.010431999961535135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,2048,0.006887110984987683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,3072,0.015847999188635085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,2560,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,2560,0.008887999587588841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,1536,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,2048,0.007153777612580194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,2560,0.015481778317027621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,2048,0.007925333248244392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,1024,0.0041582224269707995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,2048,0.013610666824711693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,1536,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,768,0.003671110918124517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,1536,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,1536,0.012493333054913415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,1024,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,1024,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,512,0.002903999967707528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,1024,0.01149511088927587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,768,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,768,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,256,0.002432000098956956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,768,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,512,0.005985777825117111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,128,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,512,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,512,0.010738666686746808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,256,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,64,0.001867555495765474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,256,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,6144,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,256,0.010565333068370819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,128,0.004943111290534337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,3584,0.01684533390733931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,6144,128,0.008267555799749162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,65536,0.10492622190051609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,64,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,6144,32,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,51200,0.08376889096366034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,65536,0.06687821944554646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,65536,0.06113244427575005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,51200,0.049350221951802574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,51200,0.05137955480151706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,12288,0.024712888730896845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,16384,0.02128622266981337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,16384,0.02017155620786879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,65536,0.11007377836439346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,51200,0.08885244528452556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,10240,0.020769778225156996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,12288,0.016787555482652452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,12288,0.01620533400111728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,8192,0.01828977796766493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,12288,0.03086044390996297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,10240,0.02032355633046892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,10240,0.026749332745869953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,6144,32,0.0019226666125986313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,7168,0.016011555989583332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,8192,0.013023110727469126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,8192,0.018012444178263348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,6144,0.014075555735164218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,8192,0.02418488926357693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,7168,0.012548444171746572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,7168,0.016765332884258695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,5120,0.012292444705963135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,16384,0.030564443932639226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,7168,0.022023111581802368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,6144,0.011223110887739392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,6144,0.014768888552983602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,4096,0.011214222345087262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,5120,0.010016000105275048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,6144,0.020933333370420668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,5120,0.014076444837782117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,3584,0.009759110709031424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,4096,0.008960000342792934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,4096,0.011706666813956367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,16384,0.037011557155185275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,4096,0.01774133410718706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,3072,0.008380444513426887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,5120,0.0192275560564465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,3584,0.010671999719407825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,2560,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,3584,0.00795555528667238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,10240,0.014816888504558139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,3584,0.016519111063745286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,3072,0.007667555577225155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,3072,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,2560,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,3072,0.015789333317014907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,2560,0.008072000410821702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,1536,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,2560,0.013751111096805997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,2048,0.007114667031500075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,2048,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,1024,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,2048,0.013034666577974955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,1536,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,768,0.00332266671790017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,1536,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,1536,0.012079999678664737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,1024,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,512,0.0028568889117903183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,1024,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,1024,0.011389333340856763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,768,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,256,0.0022817777676714794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,768,0.005865777946180767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,768,0.010627555350462595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,512,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,512,0.005434666656785541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,512,0.01070577816830741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,256,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,256,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,64,0.0018115554832749898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,256,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,32,0.0019075555933846368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,5120,128,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,2048,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,65536,0.08476889133453369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,5120,128,0.008638222184446123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,64,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,32,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,51200,0.06616710954242282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,65536,0.05962666538026598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,65536,0.058861335118611656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,16384,0.025067556235525344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,51200,0.04669955703947279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,51200,0.04762577679422167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,65536,0.10973955525292291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,16384,0.018911111685964797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,16384,0.01959733333852556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,5120,128,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,16384,0.03660088777542114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,12288,0.015976889265908133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,12288,0.015468445089128284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,10240,0.01751377847459581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,8192,0.016145777371194627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,12288,0.029417776399188574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,10240,0.01443911095460256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,10240,0.018556444181336295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,7168,0.012984000146389008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,8192,0.016036444240146212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,8192,0.012726222475369772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,8192,0.02370577719476488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,6144,0.011256888508796692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,7168,0.011380444798204633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,7168,0.0148062225845125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,7168,0.022276444567574397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,6144,0.013995556367768182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,6144,0.01036000003417333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,12288,0.020216888851589628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,51200,0.0885475542810228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,6144,0.02056711084312863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,4096,0.00903200027015474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,5120,0.009027555584907532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,5120,0.012776888906955719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,3584,0.007681777907742395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,5120,0.019174221489164565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,4096,0.010501333408885531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,4096,0.008344888687133789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,5120,128,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,3072,0.0070088886552386815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,4096,0.01717866626050737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,3584,0.007701333198282454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,3584,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,2560,0.006568000134494569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,3072,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,3584,0.016896888613700867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,10240,0.02680088910791609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,3072,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,2048,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,3072,0.015580443872345818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,2560,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,1536,0.0038844446341196695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,2560,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,2048,0.0069395556218094295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,2560,0.013422222601042854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,2048,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,1024,0.003162666741344664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,2048,0.0127697777416971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,1536,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,5120,0.010213333699438307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,1536,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,1024,0.005979555762476391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,1536,0.012132444315486483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,1024,0.0061928890645504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,512,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,1024,0.011204444699817233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,768,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,768,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,256,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,768,0.010713777608341642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,512,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,512,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,128,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,512,0.010105777945783403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,64,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,256,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,32,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,256,0.009704888694816166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,128,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,65536,0.07754044400321113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,64,0.005142222262091107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,4096,32,0.004954666727119022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,65536,0.057465778456793896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,51200,0.0588666664229499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,65536,0.058119111590915255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,4096,768,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,16384,0.021595555875036452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,51200,0.04492177897029453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,65536,0.10996355613072713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,51200,0.046727110942204796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,12288,0.01776000029510922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,51200,0.0885759989420573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,16384,0.019174221489164565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,16384,0.019136889113320243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,10240,0.015498666299713982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,16384,0.036192887359195285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,12288,0.015818667080667283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,12288,0.014910222755538093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,8192,0.01313777764638265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,4096,128,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,12288,0.02984444300333659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,10240,0.014130666851997375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,10240,0.017844445175594754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,7168,0.012028444144460889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,4096,128,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,8192,0.012055111428101858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,8192,0.015167999598715039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,6144,0.010959111154079437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,7168,0.011009777585665384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,7168,0.014124444789356656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,8192,0.02356622285313076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,5120,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,7168,0.02197155521975623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,6144,0.010298666854699453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,6144,0.013050666285885705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,4096,0.008634666601816813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,6144,0.02084622283776601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,3584,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,5120,0.01181777815024058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,5120,0.019127999742825825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,4096,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,3072,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,4096,0.016189333465364244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,3584,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,3584,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,2560,0.00611200018061532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,3584,0.016168889072206285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,3072,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,3072,0.008276444342401292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,2048,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,3072,0.01462044484085507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,2560,0.007330666813585494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,10240,0.026741334133678015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,1536,0.0041244443919923575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,2048,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,2048,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,2048,0.01272266689274046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,1024,0.0031617778456873367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,1536,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,1536,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,5120,0.009383999639087254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,1536,0.012440889245933957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,768,0.002893333426780171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,1024,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,1024,0.0057546666098965546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,4096,0.008191111187140146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,512,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,1024,0.011053333679835001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,768,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,768,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,256,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,512,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,512,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,512,0.01036711119943195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,128,0.0020702222569121257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,256,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,256,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,2560,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,256,0.009658666948477427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,64,0.0017973333597183228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,2560,0.013440888788965015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3584,32,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3584,128,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,128,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,128,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,64,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3584,32,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,51200,0.04894577794604831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,65536,0.054767111937204994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,65536,0.0568737785021464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,16384,0.01921955578856998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,65536,0.10964444610807632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,51200,0.08824266990025838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3584,768,0.011031111081441244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,12288,0.01626399987273746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,16384,0.018950222267044913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,16384,0.03571110963821411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,12288,0.015543111496501498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,10240,0.015098666151364645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,65536,0.061639110247294106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,12288,0.02920888861020406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,10240,0.01371911085314221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,10240,0.01648888819747501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,8192,0.013397333522637686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,10240,0.026741334133678015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,51200,0.04308266772164238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,8192,0.014481777946154276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,7168,0.012324444121784635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,8192,0.011759999725553723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,8192,0.023402666052182514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,6144,0.010685332947307162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,16384,0.017895110779338412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,7168,0.010515555739402771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,7168,0.013069333301650153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,7168,0.021969777014520433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,6144,0.012062222593360476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,6144,0.02012888921631707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,4096,0.007918222082985772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,5120,0.008981333010726506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,5120,0.011064888702498542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,5120,0.018568888306617737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,3584,0.007078222102589077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,4096,0.009951111343171861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,4096,0.00833422193924586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,4096,0.016533333394262526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,3072,0.006545777950021956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,3584,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,3584,0.007584888901975419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,51200,0.04587022132343716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,3584,0.0151582227812873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,2560,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,3072,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,3072,0.007819555699825287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,2048,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,3072,0.014177777700954013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,2560,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,2560,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,1536,0.003986666599909465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,2560,0.013905778527259827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,6144,0.009666666388511658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,2048,0.006909333169460297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,2048,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,1024,0.003145777723855443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,12288,0.014480888843536377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,1536,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,2048,0.01277511070171992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,768,0.002817777709828483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,1536,0.006527110934257507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,1024,0.006039999839332368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,1024,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,512,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,768,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,1024,0.011330666641394297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,256,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,768,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,768,0.011360000405046674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,512,0.005740444279379315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,128,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,512,0.01018044435315662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,256,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,256,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,64,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,256,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,32,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,3072,128,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,65536,0.05537777807977465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,128,0.008086221913496653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,64,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,3072,5120,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,65536,0.052673776944478355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,51200,0.04344977935155233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,3072,1536,0.01242844429281023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,65536,0.10975111193127102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,51200,0.0452497767077552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,512,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,16384,0.019759999381171334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,51200,0.08845244513617621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,128,0.004861333303981357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,16384,0.017664889494578045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,16384,0.01884533299340142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,12288,0.01646311084429423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,16384,0.035624000761244036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,10240,0.013816888961527081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,12288,0.015116444892353482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,3072,32,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,12288,0.01423111061255137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,8192,0.012831110921170978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,12288,0.029493331909179688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,65536,0.055997331937154136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,10240,0.015050666199790107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,10240,0.013050666285885705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,7168,0.011343110766675739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,10240,0.02637333340115017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,8192,0.010994666980372535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,8192,0.013059555656380124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,6144,0.010148444109492833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,7168,0.010311110979980892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,51200,0.04156000084347195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,7168,0.021877333521842957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,5120,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,6144,0.010987555815113915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,6144,0.009664889011118147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,4096,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,6144,0.020227554771635268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,5120,0.009019555317031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,5120,0.00831911133395301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,3584,0.007187555233637492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,4096,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,4096,0.007931555310885111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,4096,0.015848888291252982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,3072,0.006221333311663733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,3584,0.007920888562997183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,3584,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,3584,0.015268445014953613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,2560,0.0052044445441828836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,3072,0.007456888755162557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,3072,0.007126222054163615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,2048,0.004447999927732679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,3072,0.014364444547229342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,2560,0.007176000210973952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,2560,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,8192,0.02338933282428318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,2560,0.013457777599493662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,2048,0.006692444284756978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,2048,0.006225777582989798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,1024,0.003142222141226133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,2048,0.012690666649076672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,1536,0.006396444307433234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,1536,0.005983110931184556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,768,0.0028355556229750314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,1024,0.005976888868543837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,1024,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,5120,0.018139556050300598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,1024,0.011111111276679568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,512,0.002502222235004107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,768,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,768,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,256,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,768,0.010703999963071613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,512,0.00564533347884814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,512,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,128,0.0019262221952279408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,512,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,256,0.005516444643338521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,256,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,64,0.001800000046690305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2560,128,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,32,0.0018577778504954444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,128,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,7168,0.011737777955002256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,128,0.008611555728647444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,65536,0.048539555735058255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,64,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,51200,0.035331557194391884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2560,32,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,65536,0.055382221937179565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,65536,0.05091911223199633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,16384,0.016043555405404832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,65536,0.10969244771533543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,51200,0.04464888903829786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,51200,0.0404151115152571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,12288,0.013082666529549493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,51200,0.08813422256045872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,16384,0.016695110334290397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,16384,0.018393778138690524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,10240,0.01184088902340995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,12288,0.013723555538389417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,12288,0.013841778039932251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,16384,0.03569511241383023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,12288,0.028976887464523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,10240,0.014459555347760519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,10240,0.01241422196229299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,256,0.009737778041097853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2560,1536,0.0038453332251972626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,10240,0.025910221868091162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,7168,0.009739555418491364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,8192,0.010772444307804108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,8192,0.012458666331238218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,6144,0.008996444443861643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,8192,0.023367111881573994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,7168,0.010247999595271217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,7168,0.011543110840850405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,5120,0.008001777860853408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,6144,0.009462222456932068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,6144,0.00904444439543618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,6144,0.019917334119478863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,4096,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,5120,0.008799999952316284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,5120,0.008208889100286696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,5120,0.0176862229903539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,3584,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,4096,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,4096,0.015835555063353646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,8192,0.010999110837777456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,3072,0.006024888820118374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,3584,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,3584,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,3584,0.015178667174445258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,2560,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,3072,0.0076248885856734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,3072,0.006890666567616993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,3072,0.014143110977278816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,2048,0.004230222354332606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,2560,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2560,1536,0.011774222056070963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,2560,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,1536,0.0038453332251972626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,2560,0.013451555536852943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,2048,0.006666666517655055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,2048,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,1024,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,2048,0.012439111040698158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,1536,0.005962666538026597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,4096,0.008351999852392409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,1536,0.012064889073371887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,768,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,1024,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,1024,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,512,0.002507555608948072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,1024,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,768,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,768,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,256,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,768,0.010363555616802638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,512,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,512,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,512,0.010326222413116032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,256,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,256,0.005079999979999331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,64,0.0018097776919603348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,7168,0.021734222769737244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,128,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,32,0.0018560000591807896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,65536,0.0347306662135654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,128,0.00793599999613232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,64,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,2048,32,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,51200,0.027606222364637587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,65536,0.04874755607710945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,65536,0.05498488744099935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,65536,0.10933866765764023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,16384,0.012633778154850006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,51200,0.044414222240448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,51200,0.038711110750834145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,51200,0.088392886850569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,16384,0.01722311145729489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,2048,128,0.0018968888454967076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,16384,0.016414221790101793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,16384,0.035114665826161705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,10240,0.009645333720578088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,12288,0.01347555551264021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,2048,256,0.009671111073758867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,12288,0.029482665989134047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,8192,0.00870577742656072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,10240,0.01222400036123064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,10240,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,1536,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,10240,0.025791111919615004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,7168,0.008289777570300633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,8192,0.010739555789364709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,8192,0.011038222246699862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,6144,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,8192,0.022333333889643352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,7168,0.010233778092596266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,7168,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,5120,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,7168,0.02056533263789283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,6144,0.008986666798591614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,6144,0.019311111834314134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,12288,0.011300444602966309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,5120,0.009004444711738164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,4096,0.00666933341158761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,5120,0.00831911133395301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,5120,0.01720711092154185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,12288,0.013159111142158508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,3584,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,4096,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,2048,128,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,4096,0.007681777907742395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,3072,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,4096,0.015807110402319167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,3584,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,2560,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,3584,0.007119999991522894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,3072,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,3072,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,2048,0.0042133331298828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,3072,0.014455111490355598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,2560,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,2560,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,1536,0.003531555748648114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,2560,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,2048,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,2048,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,6144,0.009517333573765224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,1024,0.002995555599530538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,2048,0.013086222112178802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,1536,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,1536,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,768,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,1536,0.011744000017642975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,1024,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,512,0.0025191110455327565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,1024,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,768,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,1024,0.011400000088744693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,768,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,256,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,768,0.011110222174061669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,512,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,128,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,512,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,256,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,512,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,64,0.0018924443672100704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,256,0.01033688916100396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1536,32,0.0018231111268202465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,128,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,65536,0.02771200074089898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,128,0.008294222255547842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,64,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,32,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,51200,0.021635555558734473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,65536,0.0468444459968143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,65536,0.05430577860938179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,65536,0.1092453334066603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,51200,0.04375288883845011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,51200,0.03806666533152262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,51200,0.08727200163735284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,16384,0.01516888870133294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,16384,0.01595288846227858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,16384,0.035282668140199445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1536,256,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,10240,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1536,3584,0.01482311056719886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1536,128,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,12288,0.013156444662147097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,8192,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,12288,0.012679110798570844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,10240,0.012118221984969245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,12288,0.028127110666698877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,10240,0.0111191107167138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,7168,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,10240,0.025069332785076563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,8192,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,8192,0.009838222629494136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,6144,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,8192,0.021949332621362474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,7168,0.009378666679064432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,7168,0.009716444545321995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,5120,0.006607999818192587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,7168,0.020228443874253165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,6144,0.009336888790130615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,16384,0.010790222220950656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,4096,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,5120,0.008614222208658854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,5120,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,3584,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,5120,0.017138666576809354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,4096,0.0076248885856734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,12288,0.008488888541857401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,3072,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,3584,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,3584,0.007045333584149678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,3584,0.014818666709793938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,2560,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,3072,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,3072,0.014441778262456259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,2560,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,2560,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,2560,0.01349066694577535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,2048,0.00573600000805325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,6144,0.008887999587588841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,6144,0.018906666172875296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,1536,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,2048,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,2048,0.013066666821638743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,4096,0.007921777665615082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,1024,0.00388355553150177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,4096,0.015527110960748462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,1536,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,1536,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,768,0.003157333367400699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,1536,0.012122666670216454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,1024,0.006224888894293044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,1024,0.005471111171775394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,512,0.0028248888750871024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,3072,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,1024,0.01090666651725769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,768,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,768,0.0053022222386466126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,256,0.002199999988079071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,512,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,512,0.005141333159473208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,128,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,512,0.010210666391584609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,256,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,256,0.010050666828950247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,64,0.0018648888087934917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,128,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,1024,32,0.001964444501532449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,2048,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,128,0.008157333566082848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,65536,0.02199733257293701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,64,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,32,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,51200,0.017768889665603638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,65536,0.04617422156863742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,65536,0.1083573301633199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,51200,0.03740088807211982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,16384,0.009293333523803288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,51200,0.042897777424918286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,1024,768,0.011077333655622272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,51200,0.08698666758007473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,12288,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,1024,256,0.005336888962321811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,16384,0.01477599971824222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,16384,0.034246222840415105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,12288,0.013321778012646569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,1024,128,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,12288,0.02843466732237074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,10240,0.011053333679835001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,10240,0.011759999725553723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,8192,0.006108444597986009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,10240,0.024674667252434626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,65536,0.05414844552675883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,7168,0.006340444501903322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,8192,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,8192,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,6144,0.006401777681377199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,7168,0.009322666459613377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,7168,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,7168,0.01991911066903008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,5120,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,6144,0.009307555854320526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,6144,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,16384,0.015817777978049386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,4096,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,6144,0.0192248887485928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,5120,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,5120,0.008634666601816813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,10240,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,3584,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,5120,0.016970665918456185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,4096,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,3072,0.005480888817045424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,4096,0.007674666742483775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,3584,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,4096,0.015860444969601102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,3584,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,2560,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,3584,0.01479822231663598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,3072,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,3072,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,2048,0.004224888980388641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,3072,0.014465777410401238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,2560,0.007064888874689738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,2560,0.006606222026877933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,2560,0.013791999883121915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,2048,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,8192,0.021990221407678392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,2048,0.006214222146405115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,1024,0.0035386664999855887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,2048,0.012433778080675336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,1536,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,1536,0.005768888940413793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,768,0.0031786666562159858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,1024,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,1536,0.012453333371215396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,1024,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,512,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,1024,0.011384888655609555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,768,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,768,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,256,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,768,0.011239111423492432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,512,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,512,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,128,0.0019013333237833446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,512,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,256,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,256,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,64,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,256,0.010041777458455827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,128,0.004855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,65536,0.015631111131774057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,768,1536,0.004423111263248655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,64,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,32,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,51200,0.013385777672131857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,65536,0.04541333185301887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,65536,0.05062222149637011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,65536,0.1075893309381273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,16384,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,51200,0.03838933176464505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,51200,0.036991112762027316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,768,12288,0.012208888928095499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,12288,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,16384,0.015840000576443143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,16384,0.014720888601409065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,16384,0.03386933273739285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,10240,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,12288,0.013083555632167392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,768,128,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,12288,0.028127110666698877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,8192,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,768,128,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,10240,0.011742221812407175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,10240,0.01107911103301578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,7168,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,10240,0.024725332856178284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,8192,0.010439111126793755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,8192,0.00978311151266098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,6144,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,8192,0.021943999661339655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,7168,0.009316444396972656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,7168,0.019959999455346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,5120,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,6144,0.00903733323017756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,6144,0.008741333252853816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,51200,0.0864844454659356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,4096,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,5120,0.0086666668454806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,5120,0.008005333443482717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,3584,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,12288,0.012048888537618848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,4096,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,4096,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,3072,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,4096,0.0159608887301551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,3584,0.0074417781498697065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,3584,0.0069439998931354946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,2560,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,3584,0.015132443772421943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,3072,0.007435555259386699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,3072,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,2048,0.004861333303981357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,3072,0.014470222923490735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,2560,0.007146666447321574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,2560,0.006490666833188798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,1536,0.003559111307064692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,2560,0.013420444395807056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,2048,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,2048,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,1024,0.0029164445069101122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,1536,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,1536,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,6144,0.019194665882322524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,768,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,1536,0.011529777612951068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,1024,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,1024,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,5120,0.01686311099264357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,512,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,1024,0.010693333215183683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,768,0.0052933332820733385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,256,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,512,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,768,0.010404444403118556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,512,0.004983110974232356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,128,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,512,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,256,0.005418666534953647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,256,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,64,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,128,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,256,0.009320889082219865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,7168,0.010025777750545079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,128,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,65536,0.010420444111029306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,64,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,32,0.004534222185611725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,512,2048,0.012920888761679331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,51200,0.009290667043791877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,65536,0.046520888805389404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,65536,0.04282933473587036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,65536,0.10850666628943549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,16384,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,51200,0.03735111157099406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,51200,0.0861262215508355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,16384,0.014481777946154276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,12288,0.0053102220926019884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,16384,0.015805333852767944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,16384,0.03458311160405477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,512,32,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,12288,0.013091555900043912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,12288,0.012109333442317115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,12288,0.02775555517938402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,10240,0.011025778121418424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,10240,0.011697777443461947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,8192,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,10240,0.025011556016074285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,8192,0.010430222584141625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,512,768,0.0058764442801475525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,7168,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,8192,0.009669333696365356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,8192,0.02239377796649933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,6144,0.005881777654091517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,7168,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,7168,0.009179555707507664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,7168,0.020280000236299302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,5120,0.005346666607591841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,6144,0.008960000342792934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,512,128,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,6144,0.018570666511853535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,10240,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,4096,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,5120,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,5120,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,5120,0.01684711045689053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,4096,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,4096,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,3072,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,4096,0.015455999308162265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,3584,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,3584,0.01475822263293796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,51200,0.03494133220778571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,3072,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,3072,0.006592888798978593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,3072,0.013766222529941134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,2560,0.0069875551594628235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,2560,0.00665155549844106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,2560,0.013100444442696042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,6144,0.008977777428097194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,2048,0.005987555616431766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,1536,0.0035626664757728577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,2048,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,2048,0.013091555900043912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,1024,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,1536,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,3584,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,1536,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,768,0.0028026666906144884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,1536,0.012416000167528788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,1024,0.006220444622966979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,1024,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,512,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,3584,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,768,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,768,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,256,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,2560,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,768,0.011376000112957425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,512,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,128,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,512,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,512,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,256,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,64,0.001806222316291597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,256,0.004863111095296012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,32,0.001814222170246972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,256,128,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,128,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,65536,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,128,0.008053333395057255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,64,0.004958222309748332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,51200,0.006934222247865465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,256,32,0.0047208890318870544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,65536,0.04598133431540596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,65536,0.04254755708906385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,16384,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,65536,0.10664888885286118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,51200,0.03704799877272712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,51200,0.034244444635179304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,12288,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,51200,0.08539199829101562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,16384,0.014515555567211576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,16384,0.015821332732836407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,1024,0.010713777608341642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,10240,0.005477333234416113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,16384,0.033598221010631986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,12288,0.013220444321632385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,12288,0.0120666664507654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,8192,0.005530666559934616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,12288,0.027798222170935735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,10240,0.011746666497654386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,256,2048,0.004230222354332606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,10240,0.010830221904648675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,7168,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,8192,0.010674667027261523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,256,256,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,6144,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,8192,0.010009778042634329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,8192,0.021627555290857952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,7168,0.010045333041085137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,7168,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,5120,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,7168,0.019884443945354886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,6144,0.009325332939624786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,6144,0.008624888956546783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,4096,0.004531555705600315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,5120,0.008627555436558193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,6144,0.019276445110638935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,5120,0.007670222057236566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,4096,0.007962666451931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,4096,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,4096,0.015455111861228943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,3584,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,3072,0.004520888957712385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,3584,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,3584,0.01479822231663598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,2560,0.004527111020353106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,3072,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,3072,0.0064560001095136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,3072,0.013738666971524557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,10240,0.02455288834042019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,2560,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,2560,0.013103110922707452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,2048,0.005970666805903117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,2048,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,1536,0.003524444583389494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,2048,0.012243555651770698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,3584,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,1024,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,5120,0.01719999975628323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,1536,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,1536,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,768,0.0026319999661710528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,1536,0.011742221812407175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,1024,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,1024,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,512,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,1024,0.011069333387745751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,768,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,768,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,256,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,768,0.010392000277837118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,512,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,128,0.0018835556175973679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,512,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,2048,0.0041946665280395085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,256,0.005299555758635203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,512,0.010647111468844943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,256,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,32,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,256,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,12,128,128,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,65536,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,51200,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,128,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,16384,0.005392888767851724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,12,128,128,0.006990222467316522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,64,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,12288,0.004987555659479565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,32,0.004558222161398994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,10240,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,8192,0.00498844434817632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,65536,0.041724443435668945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,7168,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,51200,0.033599111768934466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,16384,0.014463999205165438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,12288,0.012089778151777057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,10240,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,8192,0.010002666877375709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,4096,0.00423200014564726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,7168,0.008985777695973715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,3584,0.004367111043797599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,6144,0.008719999757077958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,3072,0.004272888931963178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,5120,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,2560,0.0042257776690853965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,4096,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,2048,0.004548444516128964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,3072,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,128,2560,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,128,64,0.0018115554832749898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,1536,0.0035413333939181436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,1024,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,2560,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,768,0.002467555511328909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,2048,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,512,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,1536,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,1024,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,256,0.001800000046690305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,128,0.001790222194459703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,768,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,64,0.001656888880663448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,6144,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,32,0.001828444500764211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,256,0.004942222187916438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,64,5120,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,128,0.004515555583768421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,51200,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,32,0.004612444589535396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,16384,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,65536,0.04211466511090597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,12288,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,51200,0.0336097776889801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,10240,0.005146666533417172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,3584,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,8192,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,16384,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,7168,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,10240,0.010737777584128909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,6144,0.00452977791428566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,8192,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,5120,0.0041848888827694785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,7168,0.009009777671760982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,4096,0.004200889004601372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,6144,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,512,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,3584,0.004254222330119875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,5120,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,65536,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,2560,0.00424177779091729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,3584,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,2048,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,2560,0.006640888750553131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,1536,0.003536888708670934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,2048,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,1024,0.0028444443725877335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,12288,0.012080888781282636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,768,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,1536,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,512,0.002226666650838322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,1024,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,256,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,768,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,128,0.001812444378932317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,64,0.0017911110901170308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,512,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,32,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,12,32,3072,0.0041786668201287585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,256,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,4096,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,64,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,32,0.005017777697907554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,3072,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,16384,0.32378578186035156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,16384,0.16238310601976183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,51200,0.48853423860337997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,12288,0.2255422274271647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,16384,0.15996355480617946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,51200,0.9590755038791232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,12288,0.12147911389668782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,10240,0.18749777475992838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,12288,0.10381599929597642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,8192,0.15304356151156956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,64,64,0.004517333375083076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,10240,0.08755199776755439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,7168,0.1324675612979465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,8192,0.07176355520884196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,51200,0.4328640037112766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,8192,0.11249778005811904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,6144,0.11400266488393147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,7168,0.09460710816913181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,5120,0.0957991083463033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,4096,0.07969955603281657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,6144,0.08337155315611099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,5120,0.046895109944873385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,6144,0.054514666398366295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,3584,0.06820799907048543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,5120,0.08405955632527669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,4096,0.03980088896221585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,3072,0.05899644560284085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,4096,0.06185955471462674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,3584,0.05669599771499634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,3584,0.03533511029349433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,10240,0.12552089161343044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,12,32,128,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,2048,0.04081688986884223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,3072,0.045531554354561694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,1536,0.03141955534617106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,2560,0.02715733316209581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,2560,0.04222222169240316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,2048,0.02324177821477254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,2048,0.033992889854643084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,1536,0.018948444061809115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,1024,0.0151591118839052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,1024,0.019910222954220243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,1536,0.023736000061035156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,768,0.01750577820671929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,512,0.012503110700183444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,768,0.014450665977266101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,768,0.014074666632546319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,512,0.012411555482281579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,512,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,2560,0.049752887752321034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,256,0.010427555276287926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,128,0.005197333378924264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,3072,0.031353778309292264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,256,0.010061333576838175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,64,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,32,0.00423200014564726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,128,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,1024,0.02276977731121911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,128,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,65536,7168,0.06332621971766154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,64,0.009137777818573846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,65536,32,0.010072888599501716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,65536,256,0.007213333414660559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,51200,0.39029955863952637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,65536,0.5173866483900282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,65536,1.0881484349568684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,16384,0.2749200132158068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,16384,0.13060888979170057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,16384,0.13069689273834229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,12288,0.09770044353273179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,12288,0.20681866010030112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,12288,0.08482844299740261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,51200,0.3335546652475993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,10240,0.1741093397140503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,65536,0.4493581983778212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,8192,0.14082845052083334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,7168,0.12221511205037434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,10240,0.09943999846776326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,10240,0.06923733154932658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,8192,0.09017599953545465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,6144,0.10472444693247478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,8192,0.058145774735344775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,5120,0.08855999840630426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,7168,0.05050222078959147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,51200,0.8473982281155057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,4096,0.07154133584764269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,6144,0.06681066751480103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,5120,0.03862311111556159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,3584,0.06293511390686035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,5120,0.05918755796220568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,4096,0.05174577898449368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,3584,0.029234667619069416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,3072,0.054915553993648954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,3584,0.044249776336881846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,2560,0.045725332366095654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,3072,0.025756445195939805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,2048,0.03705777724583944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,3072,0.03700088792377048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,2560,0.02751911183198293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,2560,0.03454044461250305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,1536,0.028805332051383123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,4096,0.03231733375125461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,2048,0.028541333145565454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,2048,0.02305511136849721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,1536,0.01644533375898997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,768,0.0151591118839052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,1536,0.019890666007995605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,1024,0.013402666482660504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,512,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,7168,0.07634577486250135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,1024,0.01681777834892273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,256,0.006380444599522485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,6144,0.04440977838304308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,768,0.012862222062216865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,128,0.0032177778581778207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,512,0.011137777732478248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,256,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,64,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,512,0.011012444065676795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,32,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,256,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,51200,128,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,128,0.007335111498832703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,64,0.008341333104504479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,32,0.008835555778609382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,51200,0.2677617867787679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,65536,0.17474399672614205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,65536,0.36986933814154727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,16384,0.08922133180830215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,65536,0.1537039942211575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,51200,0.13087911076015896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,12288,0.06879822413126628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,51200,1024,0.02050933407412635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,51200,0.12275911702050103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,16384,0.04758933186531067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,16384,0.05053066545062595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,10240,0.05163822240299649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,8192,0.04344177908367581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,12288,0.03341333402527703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,10240,0.028982222080230713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,7168,0.03809066613515218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,51200,768,0.013639999760521783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,12288,0.03638133406639099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,8192,0.024549333585633173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,8192,0.03457333313094245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,5120,0.028199109766218398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,10240,0.04067911042107476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,7168,0.022222222553359136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,6144,0.019880000087949965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,7168,0.030983110268910725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,6144,0.027394667267799377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,6144,0.03261244297027588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,4096,0.023433778021070693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,5120,0.017635555730925668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,3584,0.02010311186313629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,3072,0.01795022189617157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,5120,0.0236817780468199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,2560,0.015116444892353482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,4096,0.02203111184967889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,3584,0.014386667145623101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,3584,0.018838221828142803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,2048,0.01238844460911221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,2560,0.012348444097571902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,3072,0.015857777661747403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,1536,0.009756444229020013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,2560,0.015214223000738354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,2048,0.011019555230935415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,1024,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,2048,0.013408888545301227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,1536,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,1024,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,1536,0.010457778142558204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,512,0.0041982221106688184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,1024,0.009036444127559662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,768,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,256,0.002487999904486868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,768,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,512,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,128,0.0022604444788561925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,512,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,256,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,64,0.0021093332519133887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,128,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,4096,0.015726221932305228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,32,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,128,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,3072,0.013039111263222165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,64,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,65536,0.25247110260857475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,16384,32,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,65536,0.14928355481889513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,51200,0.19765510823991564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,16384,768,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,16384,0.06508444415198432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,12288,0.05035110976960924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,65536,0.13452533880869547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,51200,0.10428533289167617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,51200,0.11230756176842584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,10240,0.04285422298643324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,16384,0.04152533411979675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,8192,0.035742223262786865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,12288,0.03182044294145372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,12288,0.02987733483314514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,7168,0.031152000029881794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,10240,0.025656888882319134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,16384,256,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,6144,0.027258666025267705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,10240,0.03660000032848782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,5120,0.023807111713621352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,8192,0.030768000417285498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,7168,0.019938665959570143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,7168,0.02686044408215417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,4096,0.018979556030697294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,6144,0.017849778135617573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,3584,0.016525333126386006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,6144,0.023952000670962866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,5120,0.01612355477280087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,3072,0.014505777094099255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,5120,0.02141600019401974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,4096,0.019671110643280875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,3584,0.013497778111033969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,2560,0.012366222010718452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,3584,0.017264000243610807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,2048,0.010564444793595208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,3072,0.014594667487674289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,2560,0.01110577748881446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,1536,0.008449777960777283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,2560,0.014255111416180929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,2048,0.00979200005531311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,1024,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,2048,0.01238222254647149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,1536,0.00923822240697013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,16384,0.044064889351526894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,1536,0.009735999835862054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,1024,0.007381333245171442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,512,0.003998222036494149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,1024,0.008098666866620382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,8192,0.021756443712446425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,768,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,256,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,768,0.006706666615274217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,512,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,128,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,512,0.006588444527652528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,256,0.006588444527652528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,64,0.0021280000607172647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,256,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,4096,0.014319111903508505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,128,0.006543111056089401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,128,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,12288,3072,0.012416000167528788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,64,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,12288,32,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,51200,0.17102932929992676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,65536,0.13914488421546087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,16384,0.05766222212049696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,768,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,51200,0.10748000277413262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,12288,0.044849776559405856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,16384,0.03994133406215244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,51200,0.09746488597657944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,16384,0.0395342210928599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,10240,0.037073777781592474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,12288,0.03042666779624091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,12288,0.028696888022952612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,8192,0.030636443032158747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,10240,0.02372177773051792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,12288,32,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,10240,0.030819555123647053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,8192,0.021156445145606995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,8192,0.02699111070897844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,6144,0.02440266642305586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,7168,0.018950222267044913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,65536,0.2156426641676161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,5120,0.020977778567208186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,7168,0.023112888137499493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,6144,0.017496888836224873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,4096,0.01774044500456916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,6144,0.02085866696304745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,5120,0.015371556083361307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,3584,0.015127999915017022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,4096,0.013786666923099093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,65536,0.1270728905995687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,5120,0.017835555805100333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,4096,0.01585244470172458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,2560,0.011350221931934357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,3584,0.014455111490355598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,3072,0.011758222348160215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,2048,0.009767110976907942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,3072,0.01311555587583118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,2560,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,2560,0.012196444802814059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,1536,0.008027555214034187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,2048,0.009299555586444007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,1024,0.006555555595291986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,1536,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,7168,0.027515555421511333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,768,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,1536,0.008630221916569604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,512,0.0038533334930737815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,1024,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,1024,0.007328888608349695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,768,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,256,0.0028897778441508612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,512,0.006852444261312485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,128,0.002444444431198968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,256,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,512,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,64,0.0021235555824306277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,32,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,10240,3072,0.013470222552617392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,128,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,3584,0.013088888592190213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,64,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,65536,0.17241511080000135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,32,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,51200,0.12811733616722956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,65536,0.09063377645280625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,65536,0.0704444448153178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,2048,0.011078221930397881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,51200,0.06918222374386258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,12288,0.034703999757766724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,16384,0.02384888960255517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,16384,0.027648000253571406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,10240,0.026740445031060114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,12288,0.018623999423450895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,8192,0.02467022173934513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,12288,0.021615111165576514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,10240,768,0.006463999963468975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,10240,0.01826755536927117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,7168,0.020608888732062448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,10240,0.024896888269318476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,8192,0.014813333749771118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,6144,0.018208000395033095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,10240,128,0.00656355544924736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,7168,0.01367377738157908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,8192,0.019967110620604623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,5120,0.015512888630231222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,7168,0.01816444430086348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,6144,0.01257777793539895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,4096,0.013020444247457715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,5120,0.011706666813956367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,6144,0.016525333126386006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,3584,0.011499555574523078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,5120,0.014081777797804939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,4096,0.010438222024175854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,3072,0.01055022246307797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,3584,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,16384,0.0440942214594947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,2560,0.008988444175985124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,4096,0.013051555388503604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,3584,0.01239999963177575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,2048,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,51200,0.05587822198867798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,1536,0.006487999939256244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,3072,0.01105955574247572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,2560,0.008064889245563084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,1024,0.004557333472702238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,2048,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,2560,0.010650667051474253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,1536,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,768,0.003504000190231535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,1024,0.00638755535085996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,1536,0.007153777612580194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,512,0.003035555490189128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,768,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,256,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,768,0.006044444524579578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,512,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,128,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,512,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,256,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,256,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,128,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,32,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,128,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,64,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,65536,0.14837066332499185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,32,0.005315555466545953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,8192,3072,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,51200,0.11552266279856364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,65536,0.08278489112854004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,2048,0.009034666750166152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,65536,0.0655768911043803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,16384,0.04011377692222595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,51200,0.06271555688646105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,12288,0.03154755632082621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,8192,1024,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,10240,0.0266266663869222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,16384,0.02572000026702881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,12288,0.017658667431937326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,8192,0.022991999983787537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,8192,64,0.001828444500764211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,10240,0.021599110629823472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,7168,0.01958666741847992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,8192,0.014135110709402295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,8192,0.02088711162408193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,6144,0.01685422162214915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,7168,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,5120,0.014774221513006421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,7168,0.016879111528396606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,6144,0.011985777980751462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,4096,0.012478222449620565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,6144,0.015820443630218506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,5120,0.011046222514576383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,5120,0.014583999911944071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,4096,0.009645333720578088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,4096,0.012429333395428128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,3072,0.009857777920034196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,3584,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,51200,0.05366577704747518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,3584,0.011729777687125735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,2560,0.008607111043400234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,16384,0.021532444490326777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,3072,0.0079644446571668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,12288,0.019900444481107924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,3072,0.010059555371602377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,1536,0.006226666685607698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,10240,0.01567911108334859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,2560,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,1024,0.004498666773239772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,2048,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,2048,0.008410666551854875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,768,0.00351555562681622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,1536,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,1536,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,512,0.002874666617976295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,1024,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,768,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,256,0.0024097777075237697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,512,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,512,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,3584,0.011015999648306104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,128,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,256,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,64,0.0018737777653667664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,256,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,128,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,7168,2048,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,64,0.0052879999081293745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,2560,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,65536,0.12156177891625299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,7168,32,0.005396444350481033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,51200,0.09557511409123738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,65536,0.0731511116027832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,1024,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,16384,0.033913777934180364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,51200,0.05701333284378052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,768,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,12288,0.027143999934196472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,51200,0.051664001411861844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,16384,0.020814221766259935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,10240,0.02318488889270359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,16384,0.023637332850032385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,12288,0.017343110508388944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,12288,0.018357333209779527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,10240,0.015200000670221118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,7168,0.01752444439464145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,8192,0.01371911085314221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,8192,0.0185199992524253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,6144,0.01569422251648373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,7168,0.012714666624863943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,7168,0.015783111254374187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,5120,0.013713777893119387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,6144,0.011736888852384357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,6144,0.01460444430510203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,4096,0.010415111151006486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,5120,0.010465777582592435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,5120,0.014099554883109199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,3584,0.009337777892748514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,4096,0.0090524446633127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,4096,0.012087110843923358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,3072,0.009135111338562435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,3584,0.008060444560315874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,3584,0.010991111397743225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,2560,0.007924444145626491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,3072,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,10240,0.022826666633288067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,3072,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,2048,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,2560,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,1536,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,2560,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,2048,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,1024,0.004239999999602635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,2048,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,1536,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,768,0.003531555748648114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,1536,0.006681777950790193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,1024,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,512,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,768,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,65536,0.06332177586025663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,256,0.002467555511328909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,768,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,512,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,128,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,256,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,64,0.0018728888697094387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,6144,128,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,256,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,32,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,128,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,6144,8192,0.01994400057527754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,65536,0.10530933406617905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,64,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,7168,128,0.0058755555914507965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,51200,0.08395644691255356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,32,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,65536,0.06697244114345975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,16384,0.03050133254792955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,65536,0.061167114310794406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,12288,0.024703111913469102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,51200,0.051298667987187706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,51200,0.04954044355286492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,10240,0.02065333392884996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,16384,0.021273778544531927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,8192,0.018332444959216647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,12288,0.016647110382715862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,12288,0.016379555066426594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,10240,0.0147724449634552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,7168,0.015916445189052157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,10240,0.018570666511853535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,1024,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,8192,0.013280889226330651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,6144,0.014081777797804939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,5120,0.012340444657537671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,8192,0.017864000466134813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,7168,0.012347555822796293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,6144,512,0.005736888696750005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,4096,0.011038222246699862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,6144,0.011039999624093374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,3584,0.009576000273227692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,5120,0.009756444229020013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,5120,0.012245333029164208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,3072,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,3584,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,4096,0.011749332977665795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,3584,0.01036800030204985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,2560,0.007450666692521837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,3072,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,3072,0.008835555778609382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,2048,0.006559111177921295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,2560,0.007270221908887227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,2560,0.008794666992293464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,16384,0.020276443825827706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,1536,0.00518844442235099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,2048,0.006986666884687211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,2048,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,1024,0.003879111260175705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,1536,0.006566222343179915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,768,0.0032951111594835916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,1536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,1024,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,512,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,1024,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,768,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,256,0.002218666589922375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,768,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,6144,0.013615111509958902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,512,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,128,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,512,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,256,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,4096,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,64,0.0018595556418100994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,5120,32,0.0019128889673286015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,256,0.005084444251325396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,65536,0.08510666423373753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,128,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,64,0.004966222163703707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,51200,0.06609333223766752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,65536,0.060573332839541964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,65536,0.05896800094180637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,51200,0.04757600029309591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,51200,0.04666666520966423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,12288,0.020270221763186984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,16384,0.019479110836982727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,16384,0.019097778532240126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,10240,0.017512889371977914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,12288,0.016134222348531086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,12288,0.015397333436542086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,8192,0.015267555912335714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,10240,0.014408000641398959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,10240,0.017514665921529133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,7168,0.01275466630856196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,5120,128,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,8192,0.012784000072214337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,6144,0.011360000405046674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,8192,0.01685422162214915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,7168,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,32,0.0052826665341854095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,7168,0.01408088869518704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,5120,0.010032888915803697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,6144,0.010558221903112201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,4096,0.009402666654851701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,6144,0.012808889150619507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,16384,0.025053333905008104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,4096,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,5120,0.01203555530971951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,4096,0.011401777466138205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,3584,0.008007999923494127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,2560,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,3584,0.009131555755933126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,3072,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,2048,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,3072,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,2560,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,5120,7168,0.016135111451148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,1536,0.004184000194072723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,2560,0.008031110796663497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,1024,0.003169777699642711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,2048,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,1536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,2048,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,768,0.00282577777074443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,1024,0.006200888918505774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,512,0.00250133333934678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,768,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,1024,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,256,0.0022159999029503926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,768,0.005538666827811136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,512,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,128,0.0021226666867733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,512,0.0053288886944452924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,256,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,3584,0.007744000189834171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,64,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,5120,0.009233777721722921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,32,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,4096,128,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,4096,3072,0.007111111448870764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,65536,0.07660177681181166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,128,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,51200,0.058760000599755176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,64,0.004960000101062986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,32,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,16384,0.022218666142887537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,65536,0.05793333053588867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,65536,0.057392001152038574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,12288,0.017846221725145977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,51200,0.04660711023542616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,51200,0.045089777972963124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,10240,0.015410666664441427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,16384,0.018244443668259513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,12288,0.015787555111779105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,1536,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,8192,0.013206221991115145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,12288,0.015107555521859063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,10240,0.014113777213626437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,7168,0.011995555626021491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,10240,0.016861332787407767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,6144,0.010934222075674268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,8192,0.012320000264379712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,7168,0.011306666665607028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,8192,0.01575288838810391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,6144,0.010039110978444418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,7168,0.01369244439734353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,4096,0.008656889200210571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,4096,256,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,6144,0.012420444852775998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,3584,0.007452444069915348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,5120,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,5120,0.011632888681358762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,4096,0.00831733312871721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,3584,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,3584,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,2560,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,3072,0.007458666960398356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,3072,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,16384,0.0192577772670322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,2560,0.007233777807818518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,2560,0.007684444387753804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,1536,0.0038942222793896994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,2048,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,2048,0.007134222322040134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,1024,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,1536,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,1536,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,5120,0.009571555587980483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,768,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,1024,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,512,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,1024,0.006199999815887875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,768,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,256,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,768,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,512,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,3072,0.006923555499977536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,128,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,512,0.005190222213665645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,4096,0.009532444179058075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,64,0.0018835556175973679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,256,0.005632888939645555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,32,0.001826666709449556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,256,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3584,128,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,65536,0.06172177526685926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,128,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,51200,0.04898933238453335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,64,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3584,32,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,65536,0.05716266896989611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,65536,0.055608888467152916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,12288,0.015524443652894763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,51200,0.0458826654487186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,51200,0.04315288861592611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,10240,0.01511111193233066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,16384,0.018927110566033255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,8192,0.013415999710559845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,12288,0.015837333268589444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,12288,0.014466666513019137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,10240,0.01751733322938283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,8192,0.011734222372372946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,8192,0.014202665951516895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,6144,0.010863111250930362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,7168,0.013593778014183044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,6144,0.009645333720578088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,6144,0.011688888900809817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,5120,0.008963555925422246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3584,2048,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,5120,0.008991111483838823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,16384,0.01924533314175076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,5120,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,3584,0.007187555233637492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,4096,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,4096,0.008993777963850234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,16384,0.017495110630989075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,3072,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,3584,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,2560,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,3584,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,7168,0.012116444607575735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,3072,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,2048,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,2560,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,1536,0.0038728887836138406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,2048,0.006693333387374878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,2560,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,2048,0.006637333167923822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,1024,0.0031679999083280563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,1536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,1536,0.00620177760720253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,768,0.0028560000161329904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,1024,0.0063031112982167145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,1024,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,4096,0.00815466625822915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,768,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,768,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,256,0.0022106667359670005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,512,0.00564177789621883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,512,0.0053075556125905775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,128,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,256,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,10240,0.013787555197874704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,64,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,256,0.004949333353175057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,32,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,128,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,3072,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,65536,0.056432002120547824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,128,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,51200,0.04336533281538221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,64,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,3072,32,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,16384,0.019633778267436557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,65536,0.056235558456844754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,65536,0.052581333451800875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,51200,0.04503822326660156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,12288,0.016234666109085083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,51200,0.04166133205095927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,10240,0.013912889692518445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,16384,0.017208000024159748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,12288,0.015113777584499784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,8192,0.012665777570671506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,12288,0.014131555954615274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,7168,0.011072888970375061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,10240,0.01276533305644989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,10240,0.014644443988800049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,8192,0.010915555887752108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,8192,0.012970666918489667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,7168,0.010320888625250923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,5120,0.00888888869020674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,7168,0.011749332977665795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,6144,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,6144,0.00999466660949919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,5120,0.008807111117574904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,3072,7168,0.010886222124099731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,5120,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,3584,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,4096,0.008010666403505538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,4096,0.007723555796676212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,3072,0.006214222146405115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,3584,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,2560,0.005474666754404704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,3584,0.007322666545708974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,3072,0.0075902218619982404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,16384,0.018734221657117207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,2048,0.004514666481150521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,2560,0.007050666544172499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,3072,512,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,1536,0.0038888889054457345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,2560,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,2048,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,1024,0.0031999999450312722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,1536,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,768,0.0028488888508743713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,6144,0.010051555931568146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,1536,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,512,0.002487999904486868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,1024,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,768,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,256,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,768,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,512,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,4096,0.007920888562997183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,128,0.002007999974820349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,512,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,256,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,128,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,32,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,128,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,65536,0.04984711276160347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,64,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,32,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,3072,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,51200,0.03538844320509169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,65536,0.05525066786342197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,16384,0.016164445214801364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,65536,0.05084977878464592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,2048,0.006549333532651265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,51200,0.04469244347678291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,12288,0.013191111385822296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2560,1024,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,51200,0.04014044337802463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,16384,0.018220444520314533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,16384,0.016791999340057373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,12288,0.013733333183659447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,8192,0.01071911139620675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,10240,0.012249777714411417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,12288,0.013931555880440606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,7168,0.00964355551534229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,10240,0.013882666826248169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,8192,0.0107004443804423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,6144,0.008944888909657797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,8192,0.012067555553383298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2560,256,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,7168,0.010013333625263637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,5120,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,7168,0.009922666682137383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,4096,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,6144,0.00945688866906696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,5120,0.008972444468074376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,6144,0.008960888617568547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,3584,0.006558222075303395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,5120,0.008299555215570662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,4096,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,3072,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,3584,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,2560,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,3584,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,3072,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,2048,0.004242666479614046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,3072,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,2560,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,10240,0.011867555479208628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,1536,0.0038240001433425476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,2560,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,2048,0.007135999699433644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,1024,0.003167111012670729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,2048,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,1536,0.0063288890653186375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,768,0.002838222309947014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,1024,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,1536,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,512,0.002463999928699599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2560,64,0.0018133332745896445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,1024,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,256,0.0021537778278191886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,768,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,512,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,128,0.0020497778637541667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,256,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,64,0.0017848888205157386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,256,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,2048,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,128,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,65536,0.03468622101677789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,4096,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,64,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,51200,0.02752711044417487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,32,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,65536,0.054745776785744556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,65536,0.04867466621928745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,51200,0.0443466669983334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,51200,0.03881244526969062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,12288,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,16384,0.01697422232892778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,16384,0.016519111063745286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,10240,0.00980355590581894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,768,0.005979555762476391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,12288,0.013230221966902414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,8192,0.008960888617568547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,10240,0.011801777614487542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,2048,512,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,10240,0.012050666742854647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,7168,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,2048,128,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,6144,0.008010666403505538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,8192,0.01070844464831882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,8192,0.010275555153687796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,5120,0.007125332951545715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,7168,0.010323555933104621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,4096,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,6144,0.009322666459613377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,7168,0.009678222239017487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,3584,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,5120,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,6144,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,3072,0.005533333453867171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,5120,0.00815288888083564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,3584,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,4096,0.007365333537260692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,16384,0.012755555411179861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,3072,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,3584,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,2048,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,3072,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,2560,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,2048,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,2560,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,1024,0.0031600000543726813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,2048,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,1536,0.006508444332414203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,12288,0.013424889081054263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,1536,0.006081777728266186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,1024,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,1024,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,512,0.0025439999169773525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,768,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,768,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,256,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,512,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,512,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,128,0.0019431112127171622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,256,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,4096,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,256,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,64,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,32,0.001791999985774358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1536,128,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,2560,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,65536,0.027779555983013574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,128,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,51200,0.021706667211320665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,1536,0.0035573335157500375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,64,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1536,32,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,16384,0.010196444061067369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,65536,0.04673155479960971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,51200,0.04382755690150791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,16384,0.015978667471143935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1536,768,0.002838222309947014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,10240,0.007687999970383114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,16384,0.01518133282661438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,12288,0.013086222112178802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,8192,0.007589333587222629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,12288,0.012437333663304647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,10240,0.011774222056070963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,7168,0.006908444480763541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,8192,0.01034577770365609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,8192,0.009938666390048133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,6144,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,7168,0.009988444546858469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,7168,0.009327111144860586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,5120,0.006308444258239534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,6144,0.009294222626421187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,6144,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,4096,0.006646222124497096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,5120,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,65536,0.054344889190461904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,5120,0.008114666574531132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,3584,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,4096,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,12288,0.008610666626029545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,3072,0.0053191110491752625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,51200,0.03775377737151252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,4096,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,3584,0.007808888951937358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,2560,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,2048,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,3584,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,3072,0.007257777783605788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,1536,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,2560,0.007109333243634965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,3072,0.006687110910813014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,1024,0.0039048890272776284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,1536,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,768,0.0032115555885765287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,1536,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,1024,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,1024,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,768,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,768,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,256,0.0021582223061058256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,512,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,256,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,256,0.004996444616052839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,128,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,128,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,64,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,128,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,32,0.0018746666610240936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,2048,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,2048,0.006220444622966979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,64,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,32,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,1024,512,0.002704000100493431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,65536,0.0540888872411516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,65536,0.04615822103288439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,16384,0.009229333036475712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,1024,512,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,51200,0.04312710960706075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,12288,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,51200,0.037189334630966187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,16384,0.015783111254374187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,16384,0.014942222171359591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,10240,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,12288,0.013080000049538083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,8192,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,12288,0.012111999922328524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,10240,0.011678222152921887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,7168,0.006734222173690796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,10240,0.011047999891969891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,8192,0.01054044481780794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,6144,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,8192,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,2560,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,65536,0.021698666943444148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,7168,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,5120,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,6144,0.009306666751702627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,6144,0.008660444782839881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,4096,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,51200,0.017581333716710407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,5120,0.008815111385451423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,3584,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,4096,0.007937777373525832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,5120,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,3072,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,4096,0.007324444750944774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,3584,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,2560,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,3072,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,3584,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,2048,0.004192000048028098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,3072,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,2560,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,2048,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,2560,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,1024,0.003563555578390757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,2048,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,1536,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,768,0.003100444459252887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,1536,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,1024,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,512,0.00253866674999396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,1024,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,7168,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,768,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,256,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,1024,10240,0.011076444553004371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,768,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,128,0.0018888889915413326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,512,0.005635555419656966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,64,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,768,128,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,256,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,65536,0.015526221858130561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,128,0.004588444613748127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,51200,0.013069333301650153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,64,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,65536,0.05143288771311442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,32,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,16384,0.007625777688291337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,65536,0.04541511005825467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,12288,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,51200,0.03852355480194092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,768,1536,0.0043333334227403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,51200,0.036779556009504534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,16384,0.015498666299713982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,16384,0.014540443817774454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,12288,0.01312711089849472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,8192,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,12288,0.012096000214417776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,7168,0.0052933332820733385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,10240,0.01075111081202825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,8192,0.010729778144094678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,6144,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,8192,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,7168,0.009713778065310584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,7168,0.009006222089131674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,6144,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,6144,0.008624888956546783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,4096,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,5120,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,5120,0.008063111040327284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,3584,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,4096,0.008033778104517195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,4096,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,3072,0.0053031109273433685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,3584,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,2560,0.005055111315515307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,3584,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,3072,0.0074346669846110875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,10240,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,2048,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,3072,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,1536,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,2560,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,10240,0.011746666497654386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,2560,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,1024,0.003148444410827425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,1536,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,2048,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,768,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,1536,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,1024,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,512,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,768,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,1024,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,256,0.0022008888837363985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,768,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,512,0.005704000178310607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,128,0.0019182221343119938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,256,0.0053102220926019884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,512,0.004965333475006951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,64,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,256,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,32,0.001810666587617662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,128,0.005330666485759947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,128,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,65536,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,64,0.004556444370084339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,51200,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,65536,0.04621600111325582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,512,32,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,16384,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,65536,0.04309689005215963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,51200,0.0373288889726003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,12288,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,512,2048,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,51200,0.03420622150103251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,10240,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,16384,0.014506666196717156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,12288,0.013117333253224691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,8192,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,12288,0.012071111136012606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,10240,0.012065778175989786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,512,5120,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,7168,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,10240,0.01072177787621816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,6144,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,768,512,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,8192,0.010632889138327705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,5120,0.005304000029961268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,7168,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,6144,0.009328888522254096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,4096,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,5120,0.00868888861603207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,6144,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,3584,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,5120,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,4096,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,3072,0.004668444395065308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,4096,0.0069822221994400024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,3584,0.007649777664078607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,2560,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,3584,0.007198221981525421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,3072,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,16384,0.01567555632856157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,3072,0.006986666884687211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,2560,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,1536,0.003540444291300244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,2560,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,2048,0.006769777586062749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,1024,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,2048,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,1536,0.0061048890153567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,768,0.002567111204067866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,1536,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,1024,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,512,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,8192,0.010021333065297868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,1024,0.005649777750174205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,768,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,256,0.002179555594921112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,7168,0.00904622260067198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,768,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,128,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,512,0.005631999837027655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,64,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,256,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,512,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,32,0.0018115554832749898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,65536,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,256,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,128,0.004603555632962121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,51200,0.0069955554273393415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,64,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,256,32,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,65536,0.045894222127066724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,51200,0.03718488746219211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,256,2048,0.004538666870858935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,51200,0.0336444444126553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,16384,0.015819556183285184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,10240,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,16384,0.014501333236694336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,12288,0.013109332985348172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,12288,0.012077333198653327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,8192,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,10240,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,7168,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,10240,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,8192,0.010479999913109673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,6144,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,8192,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,7168,0.00983911090426975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,5120,0.004578666554556953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,7168,0.009337777892748514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,6144,0.009367110828558603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,4096,0.004338666796684265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,6144,0.008390222158696916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,5120,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,3584,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,5120,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,4096,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,16384,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,4096,0.007087111473083496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,3072,0.004611555486917496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,3584,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,65536,0.041759110159344144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,12288,0.005154666801293691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,2560,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,3584,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,2048,0.004223111189073986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,3072,0.0075457774930530125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,1536,0.0035173334181308746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,2560,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,3072,0.006827555596828461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,2048,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,1024,0.0028684443483750024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,2560,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,1536,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,2048,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,1024,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,1536,0.005959110955397288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,512,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,1024,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,256,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,768,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,512,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,128,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,512,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,256,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,256,128,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,64,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,256,0.004610666798220741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,32,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,128,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,65536,0.007167110840479533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,128,0.004546666724814309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,51200,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,16384,0.0052977779673205475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,64,0.00453955555955569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,128,32,0.004599110947714912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,65536,0.042111112011803516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,51200,0.03359377715322707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,10240,0.0053013331360287136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,16384,0.014121777481502958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,12288,0.011725333001878528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,7168,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,128,768,0.002521777732504739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,10240,0.010695111420419482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,8192,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,5120,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,11,128,768,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,4096,0.004536889079544279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,7168,0.008749333520730337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,3584,0.004266666869322459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,6144,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,3072,0.0041884444653987885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,5120,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,2560,0.004223111189073986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,4096,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,2048,0.0042444442709287005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,3584,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,1536,0.003531555748648114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,2560,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,12288,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,1024,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,768,0.002473777780930201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,2048,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,512,0.0021182222084866627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,256,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,1536,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,8192,0.005134222408135732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,128,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,64,0.001779555653532346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,1024,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,768,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,32,0.001826666709449556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,65536,0.008632889224423302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,512,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,256,0.004591111093759537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,51200,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,128,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,16384,0.0052800000541739995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,64,0.004688888788223267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,12288,0.0052044445441828836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,10240,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,65536,0.04144355654716492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,51200,0.03355022271474203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,8192,0.004951999833186467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,16384,0.014488000008794995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,7168,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,12288,0.011805333197116852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,6144,0.004558222161398994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,10240,0.010762666662534079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,5120,0.004222222086456087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,4096,0.004247111164861255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,8192,0.009673777553770278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,3072,0.006689777804745569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,3584,0.004247111164861255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,7168,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,6144,0.008438222110271454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,5120,0.007676444119877285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,4096,0.007270221908887227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,3584,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,2048,0.003847111016511917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,3072,0.006508444332414203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,1536,0.003519999898142285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,2560,0.006375999914275275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,1024,0.0028328889360030494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,64,6144,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,768,0.002467555511328909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,2048,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,512,0.0021217777911159727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,1536,0.00582311095462905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,64,32,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,256,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,128,0.0018764444523387486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,1024,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,64,0.0018195555441909367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,768,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,32,0.001800000046690305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,512,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,256,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,128,0.004592000196377437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,64,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,11,32,32,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,3072,0.0041235557032956015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,11,32,2560,0.004129777765936322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,16384,0.33312710126241046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,12288,0.25332532988654244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,16384,0.1628026697370741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,51200,0.4875929090711806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,16384,0.1374666690826416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,51200,0.9938586552937826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,12288,0.1434817843967014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,12288,0.1276808844672309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,10240,0.14955643812815347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,8192,0.15415467156304255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,10240,0.08758044242858887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,7168,0.1323066684934828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,8192,0.11143199602762859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,8192,0.07146400213241577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,51200,0.40678045484754777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,7168,0.06333333253860474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,6144,0.11391022470262314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,5120,0.09583466582828098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,5120,0.04708888795640734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,6144,0.05502222312821282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,5120,0.06890844636493258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,6144,0.08548888895246719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,3584,0.06792800294028388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,3072,0.059113780657450356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,4096,0.06741866800520155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,4096,0.039191110266579524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,10240,0.18755555152893066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,3584,0.03632622294955783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,2560,0.04981955554750231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,3584,0.05427999960051643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,2048,0.04082755578888787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,3072,0.030885332160525854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,3072,0.04739288820160759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,1536,0.03130044539769491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,2560,0.03085511260562473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,1024,0.022464000516467627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,2560,0.03810133205519782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,2048,0.035953776703940496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,1536,0.01883555617597368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,1536,0.027143999934196472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,1024,0.01538044379817115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,512,0.01259022206068039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,1024,0.01919911139541202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,768,0.014513777361975776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,768,0.01685333251953125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,256,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,512,0.012730666332774691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,4096,0.07906577984491985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,512,0.012378666963842181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,7168,0.09335288736555312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,256,0.010382222632567087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,64,0.0042204442951414315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,32,0.004229333251714706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,128,0.01034311122364468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,256,0.010028444230556488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,64,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,32,0.009761778016885122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,65536,2048,0.023137778043746948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,65536,128,0.008292444050312042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,768,0.01752088963985443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,65536,0.5136924319797093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,16384,0.2516533268822564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,65536,128,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,65536,0.9702551099989148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,51200,0.38960711161295575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,51200,0.752854241265191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,16384,0.1307164430618286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,12288,0.20668978161282012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,10240,0.17319644822014701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,16384,0.10962844557232326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,12288,0.10053066412607829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,8192,0.13971111509535047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,10240,0.1201182206471761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,10240,0.06973777876959907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,12288,0.11764532989925808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,51200,0.33930667241414386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,65536,0.45391379462348086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,7168,0.12186400095621745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,8192,0.08927200237909953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,6144,0.10537511110305786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,7168,0.05195288856824239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,5120,0.08835466702779134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,7168,0.07435911231570773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,4096,0.07149510913425021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,6144,0.04494488901562161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,6144,0.0703679985470242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,5120,0.038776887787712946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,5120,0.05575111177232531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,3584,0.06268266836802165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,3072,0.05492977632416619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,4096,0.051742222574022084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,3584,0.029469334416919287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,2560,0.045977777904934354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,3072,0.03956266575389438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,3072,0.025715554753939312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,2048,0.037019554111692644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,2560,0.02533600065443251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,2560,0.03107644451989068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,1536,0.028777778148651123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,2048,0.019466666711701285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,2048,0.028575999869240656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,1024,0.020696888367335003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,1536,0.016474665866957772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,8192,0.05717244413163927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,768,0.015117333994971381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,1536,0.022499556342760723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,512,0.010623111493057676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,1024,0.016519111063745286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,768,0.013047110703256396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,256,0.0063457778758472866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,1024,0.013371555341614617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,512,0.011023999916182624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,512,0.010787555740939246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,128,0.003387555480003357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,256,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,256,0.008405333591832055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,64,0.0027351110345787476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,4096,0.03278933299912347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,51200,32,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,3584,0.04349777764744229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,128,0.007257777783605788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,64,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,32,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,65536,0.17133243878682455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,65536,0.32773423194885254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,51200,0.2614355617099338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,16384,0.08931466605928208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,65536,0.15533867147233751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,51200,0.13069866763220891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,12288,0.06950933403438993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,51200,768,0.013929777675204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,10240,0.053018665975994535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,16384,0.04731111062897576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,12288,0.04749066630999247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,16384,0.04228444563017952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,8192,0.04262755645645989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,51200,0.12489333417680527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,7168,0.03703733285268148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,10240,0.04470400015513102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,10240,0.029281778468026057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,8192,0.0244186669588089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,6144,0.03286133209864298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,8192,0.035678221119774714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,7168,0.02198666665289137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,6144,0.019740444090631273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,7168,0.03151377704408433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,4096,0.02330488959948222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,6144,0.028130667077170476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,3584,0.02033422225051456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,5120,0.02346311178472307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,4096,0.015480000111791821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,4096,0.0218551109234492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,3072,0.01789955629242791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,3584,0.014699555105633207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,3584,0.01828977796766493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,2560,0.0147733340660731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,3072,0.016399999459584553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,2048,0.0124177775449223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,2560,0.01368444412946701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,12288,0.0373342235883077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,1536,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,51200,128,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,2048,0.012834666503800286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,2048,0.013450666434235044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,1024,0.007439110842016008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,1536,0.009349333743254343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,768,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,5120,0.028057777219348486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,1536,0.011584888729784222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,512,0.004419555680619346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,1024,0.00906311141120063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,768,0.007930667036109501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,5120,0.017653332816229928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,768,0.007648000286685095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,256,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,256,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,128,0.002248888835310936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,256,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,3072,0.0173662222094006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,128,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,128,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,2560,0.014073777529928418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,32,0.0021262222694026097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,64,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,32,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,65536,0.2521786689758301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,1024,0.008024888734022776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,51200,0.19753866725497773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,65536,0.1469742192162408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,16384,512,0.007304000357786815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,16384,512,0.006483555667930179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,16384,0.06501866711510552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,51200,0.11212889353434245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,65536,0.13623466756608751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,12288,0.050310220983293324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,10240,0.042584889464908175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,16384,0.037455999188952975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,16384,0.04118577639261881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,12288,0.035240888595581055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,8192,0.03562666641341315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,51200,0.10787555244233872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,16384,64,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,12288,0.04063733418782552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,10240,0.025626666016048853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,7168,0.031207998593648274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,6144,0.02719466719362471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,10240,0.038873778449164495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,8192,0.021923555268181696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,5120,0.023711111810472276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,8192,0.030537777476840552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,6144,0.01787644459141625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,7168,0.027671110298898485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,4096,0.0196088883611891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,6144,0.02516355613867442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,5120,0.016199999385409884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,3584,0.016512889001104567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,3072,0.014665777484575907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,5120,0.020436444216304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,4096,0.014184888866212634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,3584,0.013442666994200813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,2560,0.01237155579858356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,4096,0.019648000597953796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,3072,0.014477334088749357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,2048,0.01052177780204349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,3584,0.016921778519948322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,1536,0.008289777570300633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,3072,0.015113777584499784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,2560,0.011298666397730509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,1024,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,2048,0.011178666518794166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,1536,0.008721777962313758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,2560,0.013049778011110095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,2048,0.01254666679435306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,768,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,512,0.0037742224004533556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,1024,0.007582222421964009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,1024,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,768,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,256,0.0028124445428450904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,512,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,128,0.0024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,768,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,64,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,512,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,256,0.0063999998900625445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,128,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,12288,7168,0.019908444748984445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,256,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,128,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,64,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,65536,0.21546044614579943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,32,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,51200,0.17170933882395426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,16384,0.05748977926042345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,65536,0.14229155911339655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,65536,0.12523467010921902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,12288,0.044700443744659424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,51200,0.10724711418151855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,16384,0.03509511219130622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,10240,0.03667466508017646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,51200,0.10760000016954209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,12288,0.030440890126758154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,8192,0.030488000975714788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,12288,0.03570400012864007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,12288,1536,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,10240,0.024591111474566992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,7168,0.02750933335887061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,10240,0.03124089042345683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,6144,0.024328889118300542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,8192,0.020975111259354487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,5120,0.02092622220516205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,7168,0.019345778557989333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,8192,0.02699822187423706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,6144,0.017082666357358296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,7168,0.023456000619464453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,12288,32,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,6144,0.020651555723614164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,5120,0.01577422188387977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,3584,0.015013333823945789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,5120,0.017853332890404593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,4096,0.013621333572599622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,3072,0.013391999734772576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,4096,0.016207110550668504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,3584,0.013120888835854001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,2560,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,3072,0.01216533366176817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,3584,0.014862222803963555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,2048,0.010046222143703038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,3072,0.012990222209029727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,1536,0.008016000191370646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,2048,0.008989333278603023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,16384,0.039224889543321394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,2048,0.011366222467687396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,1024,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,1536,0.00831288927131229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,1536,0.009499555660618676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,768,0.004862222406599256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,1024,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,512,0.003913777983850903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,768,0.007064888874689738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,1024,0.007430222299363878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,256,0.0028435554769304064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,768,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,512,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,4096,0.01792622274822659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,128,0.00250133333934678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,512,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,256,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,128,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,32,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,128,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,64,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,65536,0.17127911249796549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,32,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,10240,2560,0.010440000229411654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,51200,0.12886310948265925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,65536,0.09098133113649155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,16384,0.043846223089430064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,65536,0.07112177875306872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,12288,0.03456177645259433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,51200,0.05671999851862589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,16384,0.022495999932289124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,10240,0.026547556122144062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,16384,0.027833776341544256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,12288,0.020233778489960563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,12288,0.026833777626355488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,10240,0.016525333126386006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,10240,64,0.0021235555824306277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,10240,0.024298666252030268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,7168,0.020589333441522386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,8192,0.014654222461912366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,8192,0.02016622159216139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,6144,0.01811377704143524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,7168,0.013447110851605734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,5120,0.015592000550693937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,7168,0.01833866702185737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,6144,0.012678222523795234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,4096,0.013083555632167392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,6144,0.016359110673268635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,5120,0.011697777443461947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,3584,0.011561777856614856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,5120,0.01447111037042406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,3072,0.010462221999963125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,4096,0.013072888884279462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,3584,0.009834667046864828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,51200,0.0690666635831197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,2560,0.008948444492287105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,3584,0.011763555308183035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,3072,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,2048,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,3072,0.010999110837777456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,1536,0.006504000061088138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,2560,0.008282666405042013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,2560,0.009776888622177972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,8192,0.023912888434198167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,2048,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,1024,0.004701333327425851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,768,0.0035626664757728577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,1536,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,1536,0.007601777712504069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,1024,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,512,0.0028826666788922418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,768,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,1024,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,256,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,768,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,512,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,128,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,256,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,64,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,256,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,10240,2560,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,128,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,4096,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,8192,32,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,64,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,65536,0.14623467127482095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,32,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,51200,0.11407200495402019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,65536,0.08261511060926649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,16384,0.04011377692222595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,65536,0.06601511107550727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,2048,0.00927111092540953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,51200,0.06253066990110609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,12288,0.03154044349988302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,16384,0.021753778060277302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,10240,0.026564444104830425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,16384,0.02548622257179684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,12288,0.017861333158281114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,12288,0.025786666406525507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,10240,0.015638222297032673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,8192,512,0.00574311117331187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,10240,0.022304889228608873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,7168,0.01953066719902886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,8192,128,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,8192,0.014064888159434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,8192,0.019536000159051683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,7168,0.013081777426931592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,5120,0.014759999182489185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,7168,0.017183999220530193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,6144,0.011971555650234222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,6144,0.015846222639083862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,5120,0.011063999599880643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,5120,0.013801777528391944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,4096,0.009564444422721863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,3584,0.011049778097205691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,4096,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,3072,0.009662222531106737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,3584,0.008996444443861643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,51200,0.05377866824467977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,3584,0.011763555308183035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,2560,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,3072,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,3072,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,8192,0.02295733326011234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,2048,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,2560,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,1536,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,2048,0.007070221834712558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,6144,0.01697422232892778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,1536,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,2048,0.008587555752860175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,768,0.0035964445107513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,1536,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,1024,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,512,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,768,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,4096,0.012605333493815528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,1024,0.0064897777305708984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,256,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,128,0.0021235555824306277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,768,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,512,0.005962666538026597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,256,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,512,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,7168,128,0.005670222143332164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,256,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,32,0.001965333397189776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,128,0.004971555537647671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,65536,0.12126666969723171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,64,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,32,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,51200,0.09548088577058579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,65536,0.07418577538596259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,16384,0.03389599919319153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,65536,0.06326133012771606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,12288,0.026927999324268762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,51200,0.05666666560702854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,7168,2560,0.009492444495360056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,1024,0.004439110971159405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,10240,0.023200000325838726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,16384,0.020939555433061387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,16384,0.023361777265866596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,8192,0.019900444481107924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,12288,0.01715288890732659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,12288,0.02368444369898902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,7168,0.017532444662517972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,10240,0.015235554840829639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,6144,0.015697777271270752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,8192,0.013439999686347114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,10240,0.021128889587190416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,8192,0.018460444278187223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,7168,0.01258399999803967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,6144,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,7168,0.016145777371194627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,7168,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,5120,0.010515555739402771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,3584,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,5120,0.012785777449607849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,4096,0.008982222113344405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,3072,0.009020444419648912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,4096,0.012448888685968188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,3584,0.008285333712895712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,3584,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,2560,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,3072,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,3072,0.010120000276300643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,2048,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,51200,0.05149244599872165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,2560,0.0075795559419526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,1536,0.0058124446206622655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,2048,0.007002666592597961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,2048,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,1024,0.004008000095685323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,1536,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,1536,0.007249777515729268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,768,0.003510222252872255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,1024,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,4096,0.010821333361996545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,512,0.0028533333291610083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,768,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,768,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,256,0.002505777817633417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,512,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,128,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,512,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,256,0.00565155554148886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,64,0.0019031111150979996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,256,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,6144,128,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,32,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,128,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,65536,0.10541244347890218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,2560,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,32,0.005192888693677055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,51200,0.08357244729995728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,65536,0.06655555301242404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,16384,0.030474666092130873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,65536,0.06125422318776449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,6144,5120,0.013790222505728403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,51200,0.051051557064056396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,51200,0.04923911227120293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,12288,0.02460266649723053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,10240,0.020604444874657523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,16384,0.02012711101108127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,16384,0.021337777376174927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,12288,0.016558221644825406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,8192,0.018221333622932434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,10240,0.01480977733929952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,7168,0.015789333317014907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,10240,0.020626667473051283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,8192,0.013091555900043912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,6144,0.014119999276267158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,7168,0.01222400036123064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,5120,0.01219911128282547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,6144,0.011030221978823343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,7168,0.014781332678265043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,4096,0.011058666639857821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,6144,0.014112000664075216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,64,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,3584,0.009511111511124505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,5120,0.01202222208182017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,4096,0.008667555948098501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,3072,0.008546666966544257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,4096,0.011373333632946014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,3584,0.007963555554548899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,2560,0.007559110720952352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,3584,0.01073155552148819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,3072,0.007779555188284979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,3072,0.009338666995366415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,2048,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,6144,6144,0.014783999986118741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,2560,0.007471111085679796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,1536,0.005304000029961268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,2560,0.00795555528667238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,1024,0.004010666575696734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,2048,0.007038222418891058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,12288,0.022703111171722412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,1536,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,2048,0.007663110891977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,768,0.003481777591837777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,512,0.00281866660548581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,1024,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,1536,0.006783999916579988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,256,0.0024480000138282776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,768,0.006069333189063602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,8192,0.018093332648277283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,128,0.0021564445147911706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,512,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,768,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,512,0.005311111195219888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,256,0.00564444437623024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,32,0.0018257778137922287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,128,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,256,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,5120,5120,0.009645333720578088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,128,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,65536,0.0854497750600179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,64,0.0052044445441828836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,51200,0.0658551123407152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,32,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,65536,0.05901066462198893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,16384,0.02510044475396474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,65536,0.059566219647725425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,12288,0.02012977831893497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,51200,0.047488000657823354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,51200,0.04668622215588888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,16384,0.01971733404530419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,10240,0.017542221479945712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,16384,0.019401777121755812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,12288,0.016135111451148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,8192,0.015346666177113851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,12288,0.02106311089462704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,10240,0.014428445034556918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,7168,0.013074666261672974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,8192,0.012723555167516073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,6144,0.011357333097192975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,8192,0.01685688893000285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,7168,0.011740444435013665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,5120,1024,0.006219555520349079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,5120,0.010044444766309526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,7168,0.013967111706733704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,6144,0.010504888991514841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,4096,0.008948444492287105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,5120,0.009340444372759925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,6144,0.015119110544522604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,3584,0.007734222544564142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,5120,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,3072,0.007015110717879401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,4096,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,4096,0.011117333339320289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,2560,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,3584,0.009735110733244155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,3072,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,2048,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,3072,0.007886221839321984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,2560,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,1536,0.004113777644104428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,2560,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,2048,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,1024,0.0031644445326593188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,2048,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,1536,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,768,0.0028693332440323303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,1536,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,1024,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,10240,0.018386666973431904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,512,0.002499555548032125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,1024,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,256,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,768,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,512,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,5120,64,0.0018275556051068837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,512,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,256,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,64,0.0018008889423476325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,256,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,32,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,128,0.004860444615284602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,3584,0.00814577771557702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,64,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,32,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,51200,0.058706667688157826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,65536,0.05760888920889961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,65536,0.05817955732345581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,16384,0.022156443860795762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,16384,0.01921777758333418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,51200,0.04484088884459602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,12288,0.017749332719379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,16384,0.018198221921920776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,10240,0.01534577707449595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,12288,0.015847111741701763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,4096,768,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,4096,128,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,8192,0.013191999660597907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,10240,0.014104889498816596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,10240,0.017458667357762653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,7168,0.011775111158688864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,8192,0.012325333224402534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,4096,128,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,6144,0.011011555790901184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,8192,0.01573066744539473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,7168,0.011017777853541903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,5120,0.00961866643693712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,65536,0.07598844501707289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,4096,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,7168,0.013608000344700284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,6144,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,3584,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,5120,0.011364444262451597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,4096,0.00958577791849772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,3584,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,51200,0.04691822330156962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,3072,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,3072,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,2560,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,12288,0.01958755486541324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,2048,0.004973333328962326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,2560,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,2560,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,2048,0.0069919998447100324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,1536,0.0038968887594011095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,2048,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,6144,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,1024,0.0031386667655573953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,1536,0.00656088896923595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,5120,0.008988444175985124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,768,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,1536,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,4096,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,1024,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,1024,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,768,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,3072,0.006726222319735422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,256,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,768,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,512,0.005720888988839255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,128,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,512,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,256,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,64,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,32,0.0019182221343119938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,128,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,65536,0.061814222070905894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,32,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,65536,0.05481422278616163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,65536,0.05713511175579495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,51200,0.04909955461819967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3584,512,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,51200,0.046005331807666354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,16384,0.01921422282854716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,51200,0.04320266511705187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,3584,0.008942222429646386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,12288,0.01594577729701996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,16384,0.018904000520706177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,12288,0.015604444675975375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,10240,0.01514844430817498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,16384,0.01769333415561252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3584,128,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,8192,0.013386666774749756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,12288,0.017477333545684814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,7168,0.012079111403889127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,10240,0.013773333695199756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,8192,0.011401777466138205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,64,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,6144,0.01073688848151101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,7168,0.01036177741156684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,8192,0.013419555293189155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,5120,0.009031111167536842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,7168,0.012144000165992312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,6144,0.009399111072222391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,4096,0.008243555823961893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,6144,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,5120,0.008732444710201686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,3584,0.007227555745177799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,4096,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,5120,0.009559999737474654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,3072,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,2560,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,4096,0.00869333330127928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,3584,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,2048,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,3072,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,3584,0.007689778175618913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,2560,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,3072,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,1536,0.00387999994887246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,1024,0.003195555466744635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,2048,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,1536,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,2048,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,768,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,1024,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,1536,0.006227555374304454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3584,256,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,512,0.002507555608948072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,1024,0.00564444437623024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,256,0.002199999988079071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,768,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,128,0.0018942221585247253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,512,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,768,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,256,0.005713777823580637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,64,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,512,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,3072,32,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,256,0.004961777892377642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,65536,0.056461334228515625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,128,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,51200,0.04339111182424757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,32,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,65536,0.05611822340223524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,16384,0.019582221905390423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,65536,0.05235911077923245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,51200,0.04521422253714668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,51200,0.04164444406827291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,16384,0.018837332725524902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,2560,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,10240,0.013853333062595792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,12288,0.015509333875444202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,12288,0.016495111915800307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,10240,0.015263999501864115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,10240,0.013365333278973898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,8192,0.01272266689274046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,10240,0.016010666886965435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,7168,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,3072,128,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,8192,0.011050666371981302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,6144,0.010113777385817634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,8192,0.01298577752378252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,3072,64,0.004974222017659081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,7168,0.012095111111799875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,5120,0.008763555851247575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,6144,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,4096,0.007927999728255803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,5120,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,5120,0.009022222624884712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,12288,0.016157334049542744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,3584,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,4096,0.00831200016869439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,16384,0.017204445269372728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,3072,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,4096,0.007736889024575551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,2560,0.005034666508436203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,3584,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,3072,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,2048,0.004337777694066365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,3072,0.007071110937330458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,2560,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,2560,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,1536,0.0037866665257347953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,2048,0.0069439998931354946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,1024,0.0032142222755485107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,2048,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,1536,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,768,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,1536,0.006032889087994893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,1024,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,7168,0.010209778116808997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,1024,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,768,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,256,0.002157333410448498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,6144,0.011031111081441244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,512,0.005633777628342311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,512,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,128,0.0018871112002266778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,256,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,64,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,256,0.004955555415815777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,128,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,32,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2560,3584,0.007927999728255803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,128,0.004855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,64,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,51200,0.03523555397987366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,65536,0.05522933271196154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,65536,0.0505600008699629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,16384,0.01612622208065457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,51200,0.0448000000582801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,12288,0.01312622262371911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,16384,0.018201778332392376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2560,512,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,16384,0.016738666428460013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,10240,0.011737777955002256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,12288,0.014479999740918478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,768,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,8192,0.011014222270912595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,12288,0.01579466627703773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,10240,0.01298577752378252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,7168,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,10240,0.01310222182008955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,8192,0.011023999916182624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,6144,0.008958222137557136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,7168,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,8192,0.012067555553383298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,5120,0.00793511089351442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,7168,0.01012711144155926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,65536,0.049066666099760264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,6144,0.00941244430012173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,5120,0.008647999829716152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,6144,0.008981333010726506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,3584,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,5120,0.008278222547637092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,4096,0.008110221889283922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2560,32,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,3072,0.005533333453867171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,2560,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,3584,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,4096,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,3072,0.007670222057236566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,3584,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,2560,0.0069644442862934535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,3072,0.00702400008837382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,51200,0.04013777772585551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,2560,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,1024,0.0031475555151700974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,2048,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,2048,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,1536,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,768,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,1024,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,1536,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,512,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,1024,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,768,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,768,0.0052800000541739995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,512,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,128,0.0019173332386546666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,512,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,256,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,4096,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,64,0.001800000046690305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,32,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,256,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,2048,128,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,65536,0.034332444270451866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,128,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,51200,0.0273991111252043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,64,0.00489688871635331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,2048,32,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,65536,0.05470577875773112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,16384,0.012534221841229333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,65536,0.04854133394029406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,12288,0.011399110986126794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,51200,0.044327110052108765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,51200,0.038698666625552706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,16384,0.01717066599263085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,16384,0.016199111938476562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,12288,0.014431110686726041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,8192,0.008873778084913889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,12288,0.014275555809338888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,10240,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,7168,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,10240,0.012318222059143914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,8192,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,6144,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,8192,0.010267555713653564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,7168,0.010291555689440833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,5120,0.00703111125363244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,7168,0.009323555562231276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,6144,0.009687999884287516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,4096,0.006606222026877933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,6144,0.008664000365469191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,5120,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,3584,0.006015111174848344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,5120,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,4096,0.008260444634490544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,3072,0.005397333453098933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,2048,0.004321777572234471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,4096,0.00739733295308219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,2560,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,2048,1536,0.003617777592606015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,3072,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,2048,0.0042257776690853965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,10240,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,3072,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,1536,0.003547555456558863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,2560,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,2560,0.006616000086069107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,1024,0.003031111011902491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,1536,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,2048,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,768,0.0028880000528362063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,1536,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,1024,0.005976000179847081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,1024,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,512,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,256,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,768,0.005884444548024072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,512,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,128,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,512,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,256,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,64,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,128,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1536,32,0.0018159999615616268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,128,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,3584,0.008068444828192392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,65536,0.0277173337009218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,3584,0.007249777515729268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,64,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,51200,0.0218231115076277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,32,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,16384,0.010191111101044549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,65536,0.04684533344374763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,51200,0.03770577907562256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,12288,0.008457777400811514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,16384,0.0158595558669832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,16384,0.01514311134815216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,10240,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,12288,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,12288,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1536,768,0.0052826665341854095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,8192,0.0075937774446275495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,256,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,10240,0.01202755586968528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,10240,0.011386666860845355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,8192,0.010417777631017896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,6144,0.007196444604131911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,7168,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,8192,0.00980355590581894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,7168,0.009302222066455418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,5120,0.006393777827421824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,6144,0.009178666604889764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,6144,0.00871288859181934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,5120,0.00867733359336853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,4096,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,51200,0.043783999151653714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,5120,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,3584,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,4096,0.007689778175618913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,3072,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,3584,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,3584,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,3072,0.007341333561473423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,3072,0.006991110742092133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,7168,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,2048,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,2560,0.007009777757856581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,2560,0.006423111177153057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,1536,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,2048,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,2048,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,1024,0.0038613333470291565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,1536,0.006260444306664997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,1536,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,768,0.0035173334181308746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,65536,0.05444444550408257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,1024,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1536,2048,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,1024,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,768,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,256,0.0023297777192460164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,768,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,512,0.005646222167544895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,4096,0.0073626670572492815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,128,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,512,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,64,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,2560,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,128,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,65536,0.02167644434505039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,128,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,51200,0.017489777670966256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,64,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,32,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,65536,0.054024888409508594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,16384,0.009302222066455418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,65536,0.046210666497548424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,12288,0.007917332980367873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,51200,0.0370399985048506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,16384,0.015803555647532146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,16384,0.014548444085650973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,12288,0.013167999684810638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,12288,0.012088889049159156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,8192,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,1024,512,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,10240,0.011758222348160215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,10240,0.011028444601429833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,7168,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,8192,0.010383110907342698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,8192,0.009746666583749983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,6144,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,7168,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,5120,0.0064026667839950984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,1024,256,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,7168,0.009328888522254096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,6144,0.009286222358544668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,4096,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,5120,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,6144,0.00868888861603207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,3584,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,4096,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,3072,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,3584,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,51200,0.04289866818322075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,2560,0.004831111265553368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,3072,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,3072,0.0069288888739215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,2048,0.0041751112374994485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,2560,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,2560,0.006583999842405319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,1536,0.004512888689835866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,2048,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,2048,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,1024,0.00351200004418691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,1536,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,1536,0.005903111149867375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,768,0.0029644444584846497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,1024,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,1024,256,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,1024,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,768,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,768,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,256,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,512,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,4096,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,512,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,3584,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,256,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,256,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,10240,0.007012444237867991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,64,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,768,128,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,32,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,128,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,64,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,32,0.0047084444926844705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,51200,0.013048888908492194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,65536,0.04571200079388089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,65536,0.04860799842410617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,16384,0.007495111061467065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,51200,0.03870577613512675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,512,0.0025048889219760895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,51200,0.036673777633243136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,12288,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,16384,0.015843555331230164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,768,5120,0.00814400033818351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,16384,0.01443199978934394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,10240,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,8192,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,768,128,0.0021964444054497615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,12288,0.012078222301271228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,10240,0.011703111231327057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,7168,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,10240,0.011058666639857821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,8192,0.010432889064153036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,6144,0.005962666538026597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,8192,0.009899555808968013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,7168,0.009974222216341231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,5120,0.00645955569214291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,65536,0.015788444214397006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,7168,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,6144,0.009054222040706212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,4096,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,3584,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,6144,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,4096,0.008022222254011366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,3072,0.0052800000541739995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,4096,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,3584,0.007655999726719326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,3584,0.006899555524190267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,2560,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,3072,0.007252444823582967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,2048,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,3072,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,2560,0.0069546666410234236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,1536,0.0035439998739295532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,2560,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,2048,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,1024,0.0031431110368834603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,12288,0.013099555340078143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,2048,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,1536,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,768,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,1536,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,1024,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,768,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,256,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,768,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,512,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,128,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,512,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,5120,0.008784888519181145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,256,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,64,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,32,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,5120,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,512,128,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,128,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,64,0.004863111095296012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,51200,0.009278222090668148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,65536,0.046513779295815356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,65536,0.04281511240535312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,16384,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,51200,0.03716355562210083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,51200,0.03427555494838291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,12288,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,16384,0.01584533353646596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,16384,0.014484445254007975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,10240,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,12288,0.013074666261672974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,1024,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,12288,0.012015110916561551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,8192,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,10240,0.012094222009181976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,10240,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,7168,0.005387555393907759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,8192,0.010400888820489248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,256,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,6144,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,8192,0.009586667021115622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,5120,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,7168,0.009992889232105678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,6144,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,7168,0.009076444639099969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,4096,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,512,32,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,6144,0.008345777789751688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,3584,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,5120,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,4096,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,3072,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,4096,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,3584,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,2560,0.004853333450025982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,3584,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,3072,0.007576000359323289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,2048,0.00462755560874939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,3072,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,512,512,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,2560,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,1536,0.003528000166018804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,2560,0.006351111249791251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,2048,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,1536,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,2048,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,768,0.002521777732504739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,1536,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,1024,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,1024,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,512,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,65536,0.010433778166770935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,768,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,256,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,768,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,512,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,128,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,5120,0.008730666504965888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,512,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,64,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,256,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,128,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,65536,0.007898666792445714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,128,0.004570666700601578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,64,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,256,32,0.0047591109242704184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,65536,0.045886221859190196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,65536,0.04235466652446323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,16384,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,51200,0.03699822227160136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,1024,0.00286666676402092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,12288,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,16384,0.01571644511487749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,12288,0.011926222178671094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,10240,0.005543111099137201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,256,256,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,256,32,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,10240,0.010693333215183683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,10240,0.011723555624485016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,8192,0.010450666977299584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,7168,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,8192,0.009574222067991892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,6144,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,7168,0.010008888940016428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,7168,0.009039999710188972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,51200,0.03388444582621256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,5120,0.00453955555955569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,6144,0.009484444227483537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,4096,0.004294222427739037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,16384,0.014501333236694336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,5120,0.008591111335489485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,3584,0.004551110996140374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,12288,0.012744888663291931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,4096,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,3072,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,3584,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,3584,0.006660444455014334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,2560,0.004541333350870344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,8192,0.005236444373925527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,3072,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,2048,0.0042568888101312844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,2560,0.0069511110583941145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,2560,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,51200,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,1536,0.0035031110876136353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,2048,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,1024,0.0028355556229750314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,1536,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,1536,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,1024,0.005886222339338726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,6144,0.008034666379292807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,1024,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,512,0.0024782222592168385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,768,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,5120,0.007663999994595845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,768,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,512,0.005538666827811136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,512,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,128,0.0018622221218215094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,256,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,3072,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,64,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,32,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,128,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,65536,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,64,0.0045422220395671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,2048,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,32,0.004593777573770947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,51200,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,16384,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,65536,0.042097777128219604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,768,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,12288,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,10240,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,51200,0.033583998680114746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,8192,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,10,128,4096,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,12288,0.011798222031858234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,10240,0.010805333654085795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,6144,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,8192,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,5120,0.004552000098758274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,128,256,0.004587555511130227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,7168,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,4096,0.003918222255176968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,3584,0.004243555582231945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,6144,0.008716444174448649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,3072,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,5120,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,2560,0.004461333155632019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,4096,0.007320000065697565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,2048,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,3584,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,1536,0.003549333247873518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,3072,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,1024,0.0028799999919202593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,2560,0.006406222366624408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,768,0.002464888824356927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,2048,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,512,0.0021608887861172357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,1536,0.005724444571468566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,256,0.0021297778520319196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,16384,0.014461333552996317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,1024,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,128,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,128,256,0.0021466666625605687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,768,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,32,0.001773333383931054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,512,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,256,0.0048062221871482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,128,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,51200,0.007627555893527136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,64,0.004874666531880696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,64,32,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,65536,0.041477332512537636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,12288,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,51200,0.033952001068327166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,10240,0.0052915554907586836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,12288,0.011875555747085147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,8192,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,10240,0.011038222246699862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,7168,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,6144,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,8192,0.009663999908500248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,64,0.0017964444640609955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,5120,0.0042088888585567474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,4096,0.004197333421972063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,7168,0.00903200027015474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,3584,0.00423200014564726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,6144,0.008352888955010308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,65536,0.008240000241332585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,3072,0.0038933331767717996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,2560,0.004120888809363048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,5120,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,4096,0.007272000114123027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,16384,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,1536,0.00351200004418691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,3072,0.00665511108107037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,1024,0.0028799999919202593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,2560,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,768,0.0023448889454205832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,2048,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,16384,0.014125333891974555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,512,0.0021671110557185277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,1536,0.00574311117331187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,256,0.0020071110791630214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,128,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,1024,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,64,0.0017760000709030365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,768,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,32,0.0016640000459220673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,512,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,256,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,128,0.004651555584536659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,64,0.0047395556337303585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,64,7168,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,16384,0.330007102754381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,10,32,2048,0.004514666481150521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,3584,0.007255111303594377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,16384,0.16191821628146702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,51200,0.4871830940246582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,12288,0.25311822361416286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,51200,0.9898586273193359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,10240,0.18756533993615043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,12288,0.14470488495296902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,12288,0.12040444215138753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,16384,0.13538844055599636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,8192,0.1534444491068522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,10,32,32,0.004577777865860197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,10240,0.12378488646613227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,7168,0.13239999612172446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,8192,0.11217599444919163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,51200,0.4292764398786757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,10240,0.08615644772847493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,6144,0.11399288972218831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,7168,0.09652088748084174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,7168,0.06350044409434001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,5120,0.09585689173804389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,4096,0.07882933484183417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,6144,0.08045066727532281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,6144,0.05500088797675239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,5120,0.04729511009322273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,3584,0.06777333550983004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,5120,0.07359466950098674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,3072,0.05884355306625366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,4096,0.039447112215889826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,3584,0.043397333886888295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,2560,0.049626668294270836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,3584,0.05533866749869453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,2048,0.04047200083732605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,3072,0.049201776583989464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,2560,0.027172444595230952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,1536,0.031341334184010826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,8192,0.07037066751056247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,2048,0.023375110493765935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,1536,0.02714755634466807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,1024,0.022580444812774658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,1536,0.018908444378111098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,768,0.017655111021465726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,1024,0.015013333823945789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,1024,0.019645333290100098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,512,0.012396444049146442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,4096,0.06269777483410306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,768,0.016689777374267578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,768,0.015037332971890768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,256,0.007183999651008182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,512,0.012692444854312472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,256,0.010381333529949188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,128,0.005193777796294955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,3072,0.0378115541405148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,64,0.0041866666740841335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,256,0.009876444107956356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,2560,0.04103199972046746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,65536,128,0.01036800030204985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,65536,32,0.004173333446184794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,128,0.008098666866620382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,2048,0.03362577822473314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,64,0.008976889153321585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,32,0.009682666924264696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,16384,0.25084710121154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,65536,0.5130622122022841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,51200,0.3905537658267551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,51200,0.7470035552978516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,65536,0.9688533147176107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,65536,512,0.012325333224402534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,16384,0.13012533717685276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,12288,0.20664000511169434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,10240,0.17281066046820748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,16384,0.11110844214757283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,10240,0.09862489170498318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,10240,0.07014666663275824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,51200,0.34288801087273496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,8192,0.08852177858352661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,7168,0.12216444810231526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,65536,0.4349217679765489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,12288,0.09876800245708889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,6144,0.10468977689743042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,8192,0.05873777468999227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,7168,0.052101333936055504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,5120,0.08896533648173015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,6144,0.04490400022930569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,6144,0.06816444132063124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,5120,0.03924533393647935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,5120,0.06099377738104927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,3584,0.06279733445909289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,4096,0.03250844279925028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,4096,0.05168888966242472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,3072,0.05463822351561653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,3584,0.035283555587132774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,2560,0.045885334412256874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,3584,0.04506133331192864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,3072,0.026036444637510512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,8192,0.13994577195909289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,12288,0.11865777439541286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,2048,0.03707999984423319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,3072,0.039784000979529485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,1536,0.02880177895228068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,2560,0.02310666607485877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,1024,0.020241777102152508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,2048,0.019296889503796894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,1536,0.016283555163277518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,768,0.015075556106037565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,1024,0.013257777525318993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,2048,0.029147555430730183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,512,0.010670222342014313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,1024,0.016728889611032274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,7168,0.07682222127914429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,768,0.014420444766680399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,256,0.006405333264006509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,512,0.011034666664070554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,512,0.011053333679835001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,128,0.0032213332338465583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,256,0.009324444664849175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,256,0.008095111283991072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,64,0.0028693332440323303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,128,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,128,0.009541333549552495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,32,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,64,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,32,0.008700444466537898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,65536,0.32962489128112793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,1536,0.022553778356975977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,51200,2560,0.034550223085615374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,65536,0.17870488431718615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,51200,0.25986043612162274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,16384,0.08895911110772027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,51200,0.1311111052831014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,65536,0.15332621998257107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,12288,0.06915999783409967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,51200,0.12724178367190891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,10240,0.0529093345006307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,16384,0.047998223039839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,12288,0.039956443839603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,12288,0.049088888698154025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,8192,0.04349955585267809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,7168,0.037680887513690524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,10240,0.03947910997602675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,8192,0.024380443824662104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,6144,0.032024890184402466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,8192,0.03579911258485582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,7168,0.02222488820552826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,7168,0.0313297775056627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,5120,0.02796799937884013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,6144,0.01982399986849891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,4096,0.023396443989541795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,6144,0.02773333258099026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,51200,768,0.013227555486891003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,5120,0.017673777209387887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,3584,0.020317332612143624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,5120,0.024200000696712073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,4096,0.015455999308162265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,3072,0.01778133379088508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,4096,0.021167111065652635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,3584,0.014427555931939019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,2560,0.015013333823945789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,3584,0.01889600025282966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,3072,0.01643111142847273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,2048,0.01273155543539259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,16384,0.042111112011803516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,1536,0.009872000250551436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,2560,0.014816888504558139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,3072,0.017512889371977914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,1024,0.007240888973077138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,2560,0.015511999527613321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,1536,0.009242666264375051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,2048,0.013798221945762634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,768,0.005965333431959152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,1024,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,1536,0.011676444775528379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,512,0.004135999828577042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,1024,0.009031111167536842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,256,0.002536888958679305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,768,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,512,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,768,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,128,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,512,0.00646666685740153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,256,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,64,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,16384,32,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,256,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,128,0.00664533343580034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,128,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,64,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,16384,32,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,65536,0.25239822599622935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,65536,0.14863377147250706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,65536,0.13712355825636122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,16384,0.06519822279612224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,2048,0.010915555887752108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,51200,0.11211110485924615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,12288,0.050272001160515674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,16384,10240,0.0350560016102261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,16384,0.037590222226248846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,51200,0.10588533348507351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,10240,0.04281511240535312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,16384,0.04162755608558655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,8192,0.03550133440229628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,12288,0.041879110866122775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,10240,0.025825776987605627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,7168,0.030936890178256567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,10240,0.03973689013057285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,8192,0.030656887425316706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,6144,0.02706844442420536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,51200,4096,0.07185066408581205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,7168,0.019923556182119582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,5120,0.02367288867632548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,7168,0.027336888843112524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,6144,0.01796533332930671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,4096,0.01978222197956509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,6144,0.0239955551094479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,5120,0.016137777103318107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,3584,0.016545777519543965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,5120,0.02181333303451538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,51200,0.1976462205251058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,4096,0.014440889159838358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,3072,0.01460977726512485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,4096,0.01902933253182305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,3584,0.013365333278973898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,2560,0.012321777641773224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,3584,0.017483555608325534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,3072,0.012353777885437012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,2048,0.01038933296998342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,3072,0.015830222103330824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,2560,0.011376889215575324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,1536,0.00832177781396442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,2560,0.014447111222479077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,12288,0.0299217767185635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,1024,0.006665777828958299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,2048,0.010006222460005019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,1536,0.008562666674455007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,2048,0.012520888613329994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,1536,0.010496888723638324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,1024,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,512,0.003879111260175705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,768,0.0075635554061995606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,1024,0.00794488853878445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,256,0.002847111059559716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,768,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,512,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,128,0.002537777854336633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,256,0.006567111031876669
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,512,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,64,0.002116444417172008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,32,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,256,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,128,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,128,0.005396444350481033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,65536,0.2156426641676161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,32,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,12288,64,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,65536,0.14012266529930964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,51200,0.17063377963172066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,16384,0.05744000275929769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,65536,0.1286648909250895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,51200,0.10692444774839614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,12288,0.04435733291837904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,12288,768,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,51200,0.10260266727871364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,10240,0.036638223462634616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,12288,8192,0.02146222194035848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,12288,0.028608888387680054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,8192,0.030485332012176514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,16384,0.03939111034075419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,12288,0.03671822283003066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,7168,0.02741333345572154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,10240,0.03240088952912225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,8192,0.02086755633354187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,6144,0.024298666252030268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,7168,0.0188719994492001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,8192,0.026884444885783728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,5120,0.020842666427294414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,7168,0.024475556280877855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,4096,0.01753777762254079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,6144,0.01754488878779941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,3584,0.015146666102939181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,6144,0.020199111766285367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,5120,0.01551822159025404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,3072,0.013382222089502545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,5120,0.01796000036928389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,4096,0.013767110804716745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,2560,0.011583999627166323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,3584,0.012925333446926542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,4096,0.015959110524919298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,2048,0.00978488889005449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,3072,0.011993777420785693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,3584,0.015496888094478183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,1536,0.008017777568764156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,2560,0.010666666759385003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,3072,0.013811555173661975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,1024,0.006548444430033366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,2048,0.008987555901209513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,1536,0.008169777691364288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,16384,0.0355040000544654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,768,0.004836444639497333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,1024,0.007286222444640265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,1536,0.009420444567998251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,512,0.003850666599141227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,1024,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,256,0.0028533333291610083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,768,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,512,0.006755555669466655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,768,0.006679111056857639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,128,0.0024888888001441956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,256,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,64,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,256,0.005535999933878581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,128,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,10240,32,0.0022088889446523455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,64,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,32,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,65536,0.17183821731143525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,2560,0.012087110843923358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,2048,0.01093155559566286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,51200,0.12789866659376356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,65536,0.08629955848058064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,16384,0.04366933306058248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,51200,0.06873422198825412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,10240,512,0.0061919999619325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,51200,0.0582488907708062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,12288,0.03440355592303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,128,0.006573333508438534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,16384,0.026872888207435608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,16384,0.022682666778564453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,10240,0.026632888449562922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,8192,0.024295111497243244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,12288,0.0195448895295461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,10240,0.016566221912701923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,12288,0.028199109766218398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,7168,0.020578665865792167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,10240,0.02310400042268965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,8192,0.01462399959564209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,6144,0.017955554856194388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,8192,0.021328889661365088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,5120,0.01574844453069899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,7168,0.013665777941544851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,7168,0.018623999423450895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,6144,0.01257155587275823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,4096,0.014045332868893942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,6144,0.01628177695804172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,5120,0.01167377746767468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,3584,0.011582222249772815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,5120,0.014448889427714877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,4096,0.01050222251150343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,3584,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,4096,0.013776889277829064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,2560,0.008964444200197855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,3584,0.012150222228633033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,3072,0.009732444253232744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,2048,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,2560,0.008792888787057664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,2560,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,1536,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,2048,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,1024,0.00452977791428566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,2048,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,1536,0.0070471109615431885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,768,0.0037697777152061462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,1536,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,1024,0.006367110957702001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,512,0.0028560000161329904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,1024,0.006568888823191325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,768,0.0063991112013657885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,256,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,512,0.00602311102880372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,768,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,128,0.002159111201763153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,512,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,3072,0.010410666465759277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,64,0.0019057778020699818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,10240,10240,0.02467022173934513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,256,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,128,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,8192,32,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,128,0.005176888985766305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,65536,0.14678933885362414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,3072,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,64,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,8192,32,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,51200,0.11532266934712727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,16384,0.04006577862633599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,65536,0.08146044280793932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,65536,0.06656710969077216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,12288,0.03141422073046366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,51200,0.053320888015958995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,51200,0.06254311402638753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,16384,0.021457778082953557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,10240,0.026392888691690233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,8192,0.022916444473796423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,12288,0.01789955629242791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,16384,0.02583644456333584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,12288,0.02625511089960734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,10240,0.015832000308566622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,7168,0.01939288940694597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,6144,0.016890666551060148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,10240,0.022321777211295232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,8192,0.0141004439857271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,7168,0.013099555340078143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,8192,0.020969778299331665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,6144,0.012061333490742577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,4096,0.012784000072214337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,6144,0.01567911108334859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,5120,0.010890666809346942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,65536,0.0714320010609097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,3584,0.011004444625642566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,5120,0.013849777479966482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,4096,0.009671111073758867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,3072,0.009722666607962714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,3584,0.008622222476535374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,4096,0.013071111506885953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,2560,0.008641777767075432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,3584,0.011719110939237805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,3072,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,2048,0.007245333658324347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,2560,0.0073724447025193115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,2560,0.009966221948464712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,1536,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,2048,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,2048,0.008630221916569604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,1024,0.004212444441186057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,1536,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,1536,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,768,0.003294222263826264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,5120,0.014655111564530266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,1024,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,512,0.0028968888024489083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,1024,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,7168,0.01754488878779941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,256,0.002480888946188821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,8192,256,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,768,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,128,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,512,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,768,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,256,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,64,0.0018986666368113623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,512,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,7168,32,0.002060444404681524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,7168,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,256,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,128,0.005061333378156026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,65536,0.12125421894921197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,64,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,32,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,51200,0.09529955519570245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,7168,3072,0.010851555400424533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,16384,0.03392444385422601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,65536,0.0636133352915446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,65536,0.07369066609276666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,12288,0.026833777626355488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,10240,0.023073777556419373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,51200,0.05150489012400309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,51200,0.05656977494557699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,12288,0.01699022286468082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,8192,0.019846222466892667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,16384,0.023581332630581323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,7168,0.017518222332000732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,12288,0.023738667368888855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,10240,0.015212444795502557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,6144,0.015468445089128284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,8192,0.013726222018400827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,10240,0.02019466625319587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,5120,0.013746666411558786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,7168,0.01258488910065757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,8192,0.019578667150603402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,4096,0.01069777790043089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,7168,0.01623644431432088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,6144,0.015786666009161208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,5120,0.010416000253624385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,4096,0.008926221893893348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,5120,0.012842666771676807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,3072,0.009222221871217092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,4096,0.012437333663304647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,3584,0.008047999607192146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,2560,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,3584,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,2048,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,2560,0.007429333196745978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,3072,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,1536,0.005701333284378052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,2560,0.009372444616423713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,2048,0.006973333656787872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,1024,0.004411555412742826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,2048,0.007720000214046902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,1536,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,16384,0.021112889051437378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,1536,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,1024,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,512,0.0028560000161329904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,1024,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,768,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,256,0.002502222235004107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,768,0.0058142224119769196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,512,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,6144,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,128,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,512,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,256,0.00563822231358952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,32,0.0018737777653667664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,256,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,128,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,65536,0.1053217781914605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,64,0.005152889009979037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,6144,3072,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,32,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,51200,0.08384178082148235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,65536,0.06658577919006348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,16384,0.030235555436876085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,65536,0.0609146687719557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,51200,0.05121511220932007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,51200,0.049343112442228526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,12288,0.02439022229777442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,768,0.0033191111352708605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,16384,0.02017866737312741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,10240,0.020587555236286588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,16384,0.02122488949033949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,8192,0.0180711117055681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,12288,0.0166284441947937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,10240,0.014846222268210517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,7168,0.015857777661747403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,10240,0.01976977785428365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,8192,0.013156444662147097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,6144,0.014137778017255994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,8192,0.01828266680240631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,6144,3584,0.009343999955389235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,7168,0.012256888879670037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,5120,0.012334222594896952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,7168,0.016014221641752455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,6144,0.011284444067213269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,4096,0.011032000184059143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,6144,128,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,6144,0.015150222513410779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,3584,0.009515555368529426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,5120,0.009714666340086196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,5120,0.012717333104875354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,4096,0.008560889297061497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,3584,0.008209777375062307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,4096,0.012095111111799875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,2560,0.0074240002367231585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,3584,0.011010666688283285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,3072,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,2048,0.006568000134494569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,3072,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,2560,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,1536,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,2048,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,1024,0.0038737778862317405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,1536,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,1536,0.006654222392373615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,12288,0.024337778488794964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,768,0.003551111039188173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,1024,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,512,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,1024,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,768,0.006176888942718506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,256,0.002307555534773403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,512,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,512,0.005392000079154968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,128,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,256,0.005195555587609609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,3072,0.00846400029129452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,5120,32,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,5120,128,0.00563733321097162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,128,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,65536,0.08420533604092068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,64,0.004987555659479565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,32,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,2560,0.008752889103359645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,51200,0.06567111280229357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,16384,0.025026667449209426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,65536,0.059128887123531766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,2048,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,12288,0.020136000381575692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,51200,0.047525333033667676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,51200,0.04659733176231384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,16384,0.019499555230140686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,16384,0.019110222657521565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,12288,0.015965332587560017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,8192,0.015014222926563688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,12288,0.021288888321982488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,10240,0.014497778481907315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,7168,0.01274577776590983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,5120,768,0.005674666828579373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,10240,0.018208000395033095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,8192,0.012753778033786349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,6144,0.011335111326641507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,5120,0.009975999593734741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,7168,0.011373333632946014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,8192,0.016159999701711867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,4096,0.009226666556464301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,6144,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,5120,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,6144,0.013302221894264221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,3584,0.007891555627187094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,4096,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,3072,0.00703288863102595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,4096,0.01035199976629681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,3584,0.007754666937722101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,2560,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,3584,0.009712888962692684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,3072,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,2048,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,65536,0.05880977710088094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,3072,0.008623110751310984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,2560,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,1536,0.00397511116332478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,2560,0.008216888540320927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,10240,0.01751377847459581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,2048,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,1024,0.003134222287270758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,2048,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,768,0.0028293333533737394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,1024,0.006217777729034424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,1536,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,512,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,1024,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,768,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,768,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,7168,0.015418666932317944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,512,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,128,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,256,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,5120,0.012309333516491784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,256,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,64,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,32,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,128,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,65536,0.07532799906200834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,64,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,4096,32,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,51200,0.058766219351026744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,65536,0.057886222998301186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,65536,0.057909336354997426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,16384,0.02208711041344537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,1536,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,51200,0.04480088750521342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,12288,0.017616889543003507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,16384,0.01923022170861562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,10240,0.015248000621795654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,12288,0.0158924443854226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,16384,0.018055111169815063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,4096,256,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,8192,0.013220444321632385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,12288,0.02051822178893619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,4096,512,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,7168,0.012051555845472546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,10240,0.014120888378885059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,6144,0.011056000159846412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,8192,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,10240,0.019089778264363606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,5120,0.009551111194822524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,7168,0.011077333655622272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,8192,0.015499555402331881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,4096,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,7168,0.015154666370815702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,3584,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,6144,0.012633778154850006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,4096,0.008301333420806462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,3072,0.006868444383144379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,4096,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,3584,0.007816889219813878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,2560,0.005983110931184556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,3584,0.008629333641793992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,3072,0.007631999750932057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,2048,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,3072,0.008127999802430471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,2560,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,1536,0.003826666623353958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,51200,0.04664533336957296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,2560,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,2048,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,1024,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,2048,0.007196444604131911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,1536,0.0063991112013657885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,1024,0.006062222023804982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,1536,0.0064995553758409284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,512,0.002467555511328909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,1024,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,768,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,256,0.0023742222951518162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,768,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,512,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,6144,0.009679999616410997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,128,0.0021626665774318907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,512,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,256,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,64,0.0017991111510329777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,5120,0.00904888908068339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,32,0.0017955555684036678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,5120,0.012049777640236748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3584,128,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,65536,0.06159733401404487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,128,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,64,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,32,0.004967999955018361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,65536,0.05704355239868164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,65536,0.054568890068266124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,51200,0.04583022329542372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,51200,0.043140444490644664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,12288,0.015476443701320224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,16384,0.018927110566033255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,16384,0.01754044493039449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3584,768,0.0028479999552170434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,12288,0.015493333339691162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,12288,0.01833599971400367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,8192,0.013434666726324292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,10240,0.013766222529941134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,10240,0.015568888849682279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,7168,0.01202666676706738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,8192,0.011658666862381829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,8192,0.013532444834709167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,6144,0.010764444039927589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,7168,0.010504888991514841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3584,256,0.004980444494220945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,5120,0.008981333010726506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,7168,0.01260088880856832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,51200,0.04888977938228183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,6144,0.009367999931176504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,4096,0.008266666697131263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,5120,0.008984888593355814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,16384,0.019066666563351948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,3584,0.007260444263617198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,5120,0.010313777459992303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,4096,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,3072,0.006619555668698416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,4096,0.008416888614495596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,3584,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,2560,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,3584,0.008253333469231924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,10240,0.015152888165579902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,2048,0.004543999830881755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,2560,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,1536,0.003871110992299186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,2560,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,2048,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,1024,0.0032106666929192017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,2048,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,1536,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,768,0.00282577777074443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,1024,0.006027555714050929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,1024,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,512,0.002505777817633417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,768,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,768,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,256,0.0021395554973019496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,512,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,512,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,6144,0.012096000214417776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,128,0.001875555556681421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,256,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,64,0.001790222194459703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,256,0.005038222091065513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,3072,32,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,128,0.005646222167544895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,128,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,65536,0.05597866906060112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,3072,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,64,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,51200,0.04313155677583483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,32,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,16384,0.019549333386951022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,65536,0.056032001972198486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,65536,0.05249955587916904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,12288,0.016188444362746347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,51200,0.045202665858798556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,51200,0.0414648883872562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,3072,1536,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,16384,0.01870666609870063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,10240,0.013791110780504016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,8192,0.012659555508030785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,16384,0.01720533271630605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,12288,0.014888000157144336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,10240,0.012981333666377597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,7168,0.011081777513027191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,12288,0.017550221747822232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,6144,0.010055111514197456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,8192,0.011407111254003314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,10240,0.014507555299335055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,5120,0.008769777913888296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,7168,0.010071111222108206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,8192,0.013526221944226159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,4096,0.007816000117195977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,7168,0.012720000412729053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,6144,0.009675555759006077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,5120,0.008970666262838576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,3584,0.0069066666894488865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,6144,0.010992888775136737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,3072,0.00619911112719112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,4096,0.008023999631404877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,5120,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,2560,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,3584,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,4096,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,2048,0.004569777597983678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,3072,0.007290666302045186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,3072,3072,0.007365333537260692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,3584,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,1536,0.003710222327046924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,2560,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,1024,0.003157333367400699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,3072,0.0070257774657673305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,2048,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,768,0.002871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,1536,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,2048,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,1024,0.006002666635645761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,512,0.0024622221373849446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,1536,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,256,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,768,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,512,0.005620444400442972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,768,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,128,0.0019502221710152095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,512,0.005196444276306364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,256,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,64,0.001793777777089013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2560,32,0.0018133332745896445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2560,128,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,256,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,65536,0.04892266790072123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,128,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,51200,0.035134222772386335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,64,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,32,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,16384,0.015951999359660678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,65536,0.05513688921928406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,65536,0.05012089014053345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,12288,0.013218666944238873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,51200,0.04486666785346138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,51200,0.04013244311014811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,10240,0.011742221812407175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,16384,0.016812443733215332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,12288,0.013905778527259827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,8192,0.010775110787815519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,2560,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,12288,0.015740444262822468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,7168,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,10240,0.012255111502276527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,10240,0.013447999954223633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,8192,0.01072177787621816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,6144,0.008961777720186446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,7168,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,8192,0.011532444092962476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,5120,0.008024888734022776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,7168,0.010256888965765635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,6144,0.009726222190592024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,4096,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,6144,0.008991999758614434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,3584,0.0064026667839950984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,4096,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,4096,0.007684444387753804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,3072,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,3584,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,3584,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,2560,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,3072,0.007348444726732042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,16384,0.018177777528762817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,3072,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,2048,0.004233777936961916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,2560,0.0069875551594628235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,1536,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,2560,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,2048,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,1024,0.0031617778456873367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,1536,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,2048,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,768,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2560,1024,0.005330666485759947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,1536,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,512,0.0025111111915773815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,1024,0.005995555470387141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,256,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,1024,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,768,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,128,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,512,0.005674666828579373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,5120,0.008611555728647444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,768,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,64,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,2048,32,0.0018533333722088074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,5120,0.00816533300611708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,512,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,128,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,256,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,65536,0.03476266728507148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,51200,0.027274666561020747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,128,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,64,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,16384,0.01256711118751102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,65536,0.05481777919663323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,12288,0.011361777782440186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,51200,0.04422755704985725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,16384,0.017536888519922893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,51200,0.03867377837498983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,10240,0.009859555297427708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,16384,0.016356445021099515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,12288,0.013079110946920184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,8192,0.008940444224410588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,12288,0.014427555931939019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,10240,0.012095111111799875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,7168,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,10240,0.012591111163298288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,8192,0.010760889285140567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,6144,0.008007999923494127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,8192,0.010030222435792288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,7168,0.010337777435779572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,5120,0.007079111205206976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,7168,0.009367110828558603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,6144,0.009304888546466827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,4096,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,6144,0.008683555656009251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,5120,0.008838222258620793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,2048,256,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,3584,0.006026666611433029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,4096,0.008336000144481659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,3072,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,4096,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,3584,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,2560,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,3584,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,2048,32,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,3072,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,2048,0.00425155543618732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,65536,0.04843644301096598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,1536,0.003569777641031477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,2560,0.00699733363257514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,3072,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,1024,0.0031395554542541504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,2048,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,2560,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,768,0.002879111096262932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,1536,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,2048,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,512,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,1024,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,768,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,256,0.002135111019015312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,512,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,768,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,128,0.001863111017478837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,512,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,256,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,64,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,256,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1536,128,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1536,32,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,5120,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,128,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,64,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,51200,0.02164266672399309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,65536,0.05414844552675883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,65536,0.046889778640535146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,16384,0.010652444428867765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,51200,0.043848001294665866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,51200,0.037492444117863975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,12288,0.008447111480765873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,16384,0.015942222542232938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,16384,0.01567555632856157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,1536,0.005640000104904175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,10240,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,1024,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,12288,0.01310400002532535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,8192,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,12288,0.012607110871209038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,7168,0.0070471109615431885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,10240,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,6144,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,10240,0.011078221930397881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,7168,0.010053333308961656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,8192,0.00981955561372969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,5120,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,6144,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,65536,0.02769244379467434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,6144,0.008960000342792934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,5120,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,3584,0.005958222266700532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,5120,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,4096,0.00813244448767768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,3072,0.005347555710209741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,4096,0.007505777809354994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,3584,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,2560,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,3072,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,3584,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,2048,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,3072,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,2560,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,1536,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,2560,0.00656088896923595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,2048,0.006743999818960826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,1024,0.003662222375472387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,2048,0.006254222244024277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,1536,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,1536,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,8192,0.01073333372672399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,1024,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,512,0.0027315556589100095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,1024,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,768,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,256,0.002201777779393726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,512,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,768,0.0052720002002186244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,512,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,128,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,256,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,64,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,1024,128,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,32,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,128,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,65536,0.021695110532972548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,64,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,32,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,51200,0.017460443907313876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,65536,0.05405333307054308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,65536,0.0461715559164683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,768,0.0033333332588275275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,16384,0.009240888886981541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,51200,0.04349689020050896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,51200,0.037309332026375666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,12288,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1024,7168,0.009304000271691216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,16384,0.015795555379655626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,1024,4096,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,10240,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,8192,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,16384,0.014456889695591398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,12288,0.013189333180586496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,7168,0.006305777778228124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,12288,0.012303999728626676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,1536,32,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,8192,0.01036000003417333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,6144,0.006221333311663733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,10240,0.011040888726711273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,5120,0.006471111128727595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,8192,0.009648000200589498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,7168,0.00998399986161126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,4096,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,6144,0.00921688891119427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,7168,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,3584,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,5120,0.008616000413894653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,6144,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,4096,0.00793511089351442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,5120,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,3584,0.0075662218862109715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,4096,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,2560,0.004853333450025982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,3584,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,3072,0.007625777688291337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,2048,0.0042204442951414315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,3072,0.00686488880051507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,2560,0.006917333437336816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,1536,0.004457777573002709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,2560,0.006562666760550604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,2048,0.006994666324721442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,1024,0.003552000141806073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,1536,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,2048,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,768,0.003216888962520493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,1536,0.005828444328573015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,512,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,1024,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,10240,0.01202755586968528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,768,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,256,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,512,0.005662222289376789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,128,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,256,0.005509333478079901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,256,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,64,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,128,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,3072,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,768,32,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,128,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,65536,0.015535111228624979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,51200,0.013108444710572561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,65536,0.051411555873023145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,32,0.004805333498451444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,65536,0.045194665590922035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,16384,0.007644444704055786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,51200,0.036631110641691424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,51200,0.038415110773510404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,12288,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,16384,0.015830222103330824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,10240,0.005922666854328579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,1024,0.00618577789929178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,12288,0.013087111214796701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,8192,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,768,768,0.005786666853560342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,12288,0.01219466659757826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,10240,0.01203555530971951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,7168,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,10240,0.01107377807299296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,8192,0.010398222340477837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,6144,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,512,0.005195555587609609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,7168,0.010047111246320937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,8192,0.009680888719028896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,7168,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,6144,0.009344889058007134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,4096,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,5120,0.008449777960777283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,6144,0.008608000146018134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,3584,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,5120,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,4096,0.007865777446164025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,4096,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,768,64,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,3584,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,2560,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,3584,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,3072,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,2048,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,3072,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,2560,0.00699733363257514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,1536,0.003677333394686381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,2560,0.006463999963468975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,16384,0.014444443914625378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,2048,0.0063991112013657885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,1024,0.0029884444342719186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,2048,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,768,0.0028275555620590844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,1536,0.005640888793600931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,1024,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,512,0.002487999904486868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,1024,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,5120,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,768,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,256,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,768,0.005395555661784277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,512,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,128,0.0018560000591807896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,256,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,64,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,256,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,32,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,512,3072,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,65536,0.010646222366227044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,128,0.004644444419278039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,51200,0.009321778184837764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,64,0.004624888714816836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,65536,0.046634667449527316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,32,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,16384,0.005283555636803309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,65536,0.04212799999448988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,12288,0.005127111242877113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,51200,0.03425866696569655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,16384,0.015788444214397006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,10240,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,16384,0.014442667365074158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,512,1536,0.006467555546098285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,12288,0.013088000317414602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,8192,0.00544088871942626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,12288,0.012065778175989786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,10240,0.011772444678677453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,10240,0.010792888700962067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,8192,0.010704889065689512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,6144,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,8192,0.009711110757456886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,7168,0.009707555174827576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,512,512,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,5120,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,6144,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,6144,0.008369777765538957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,5120,0.008626666333940294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,3584,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,5120,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,4096,0.00794222205877304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,3072,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,4096,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,3584,0.007518221934636434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,2560,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,3584,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,3072,0.0075839997993575195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,2048,0.004599110947714912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,3072,0.006704000135262807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,2560,0.007096000015735626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,1536,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,2560,0.006301333506902059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,2048,0.00657155571712388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,1024,0.00286666676402092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,2048,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,1536,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,768,0.00258577780591117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,1536,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,1024,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,512,0.002505777817633417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,7168,0.00944266633854972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,768,0.006079111248254776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,256,0.0021608887861172357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,768,0.005230222311284807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,512,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,128,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,512,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,256,0.005299555758635203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,64,0.001785777716173066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,256,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,32,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,128,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,65536,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,256,51200,0.03728888763321771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,128,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,64,0.004651555584536659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,51200,0.00703111125363244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,32,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,65536,0.04602488875389099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,16384,0.0057751110030545135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,65536,0.04177688890033298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,51200,0.03735733363363478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,7168,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,12288,0.005200888961553574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,51200,0.03360266817940606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,16384,0.015595555305480957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,10240,0.0054551110499435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,16384,0.014479999740918478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,12288,0.013222222526868185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,8192,0.0053093334039052325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,256,1024,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,12288,0.01181688904762268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,10240,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,7168,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,6144,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,256,4096,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,10240,0.01070577816830741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,5120,0.004591111093759537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,8192,0.009709333380063375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,6144,0.009378666679064432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,7168,0.009015999734401703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,4096,0.004555555681387584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,6144,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,5120,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,3584,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,5120,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,3072,0.004514666481150521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,3584,0.007559110720952352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,3584,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,3072,0.007255111303594377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,2560,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,3072,0.00664444433318244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,2048,0.004187555362780889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,2560,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,2560,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,2048,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,1536,0.003510222252872255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,2048,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,1024,0.0028115556471877625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,1536,0.00647288892004225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,8192,0.010667555862002902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,768,0.0025315555847353404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,1024,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,1536,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,1024,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,768,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,256,0.0019724443554878235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,768,0.005384888913896348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,512,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,128,0.0018257778137922287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,512,0.004951111144489712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,4096,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,4096,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,32,0.0018195555441909367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,256,0.004940444396601783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,128,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,65536,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,51200,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,128,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,16384,0.005427555491526921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,64,0.004576888763242298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,12288,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,128,32,0.004553777890072929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,65536,0.04157244496875339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,10240,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,8192,0.005168000029193031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,51200,0.03389777739842733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,16384,0.014478221535682678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,7168,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,6144,0.004586666822433472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,12288,0.011797332929240333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,5120,0.0042160000238153665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,10240,0.01074399964676963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,4096,0.004187555362780889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,8192,0.009364444348547194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,7168,0.010009778042634329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,3584,0.004276444514592488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,7168,0.009331555830107795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,3072,0.004222222086456087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,6144,0.008584000170230865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,2560,0.004268444246715969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,2048,0.003945777813593547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,5120,0.007690666450394525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,1536,0.003530666646030214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,3584,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,1024,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,3072,0.00666844430896971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,9,128,256,0.005631111148330901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,768,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,2560,0.006365333166387346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,512,0.0025191110455327565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,256,0.002130666747689247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,1536,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,1024,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,128,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,768,0.005333333379692501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,512,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,64,0.0017946666727463405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,64,32,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,256,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,65536,0.008260444634490544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,128,0.004594666676388847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,64,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,51200,0.00739822205570009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,32,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,16384,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,12288,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,65536,0.041532443629370794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,128,512,0.002448888909485605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,51200,0.03391911254988776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,16384,0.014429334137174817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,8192,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,12288,0.011759999725553723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,7168,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,4096,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,6144,0.004564444637960858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,5120,0.004562666846646203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,10240,0.010660444696744284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,4096,0.0038888889054457345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,8192,0.009848000274764167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,64,2048,0.00598933340774642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,3584,0.004685333205593957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,6144,0.008358221915033128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,5120,0.00793333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,2560,0.004358222087224324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,4096,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,2048,0.0038977778620190094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,3584,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,1536,0.003216888962520493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,3072,0.006629333313968446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,1024,0.0028151110228565005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,2560,0.006534222099516127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,768,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,2048,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,512,0.0021768889079491296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,1536,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,256,0.001820444439848264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,1024,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,128,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,64,0.001548444438311789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,768,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,32,0.0017911110901170308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,512,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,256,0.004967111266321606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,128,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,64,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,32,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,9,32,7168,0.008984888593355814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,3072,0.004553777890072929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,16384,0.3390124373965793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,16384,0.162191112836202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,51200,0.4875217543707953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,51200,1.0128542582194011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,12288,0.24019821484883627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,9,32,10240,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,12288,0.14011111524369982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,10240,0.20161778397030303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,12288,0.1038515567779541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,10240,0.14645866552988687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,8192,0.15299199687110052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,10240,0.08739911185370551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,16384,0.17430310779147676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,8192,0.14026489522722033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,7168,0.1321128871705797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,16384,0.1570542256037394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,8192,0.07163466347588433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,10240,0.11267111036512588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,12288,0.13186666700575086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,6144,0.11406044165293376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,51200,0.4079742166731093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,8192,0.09369688563876682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,7168,0.0634382234679328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,6144,0.08255555232365926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,5120,0.09568355480829875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,7168,0.11750045087602402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,6144,0.05439466569158766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,4096,0.07892000012927584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,5120,0.07192444139056735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,7168,0.08327288760079278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,5120,0.04701333244641622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,3584,0.06780533658133613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,4096,0.06182399723264906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,4096,0.03923199905289544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,5120,0.06336177719963922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,3072,0.0588231086730957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,3584,0.05644889010323418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,3584,0.04384266667895847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,4096,0.05380444394217598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,2560,0.04981688989533318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,3072,0.04888711041874356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,3584,0.04865688747829861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,2560,0.041294223732418485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,2048,0.04036088784535726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,3072,0.04342666599485609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,2560,0.03358488943841722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,2560,0.038223998414145574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,1536,0.03176444437768724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,2048,0.023687111006842718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,2048,0.03443110982577006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,2048,0.03403822249836392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,1536,0.026732444763183594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,1024,0.022273777259720698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,1536,0.028767999675538804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,6144,0.07337422503365411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,768,0.017861333158281114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,1024,0.019550222489568923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,1024,0.017496888836224873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,1024,0.02275377843115065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,512,0.012614222036467658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,3072,0.030869334936141968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,768,0.016606221596399944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,768,0.021116443806224402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,256,0.007190222541491191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,512,0.01275111072593265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,512,0.016343999240133498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,51200,0.5043919881184896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,512,0.012723555167516073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,128,0.005105777747101254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,256,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,256,0.009742221898502773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,64,0.004231111043029361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,256,0.014273777604103088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,1536,0.019322666856977675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,65536,32,0.0042151109211974675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,128,0.008160889148712158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,128,0.010509333676762052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,64,0.009196444518036312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,65536,32,0.010010666317409938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,768,0.015104000767072042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,65536,0.5142818027072483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,65536,0.9684036042955187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,51200,0.3893786536322699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,128,0.01034400032626258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,16384,0.13023466534084743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,16384,0.24969598982069227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,65536,0.44751469294230145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,12288,0.1911111142900255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,51200,0.31734222835964626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,16384,0.12764178382025823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,12288,0.11584444840749104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,10240,0.160424894756741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,12288,0.09381333324644302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,10240,0.09935466448465984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,10240,0.07201600074768066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,8192,0.13942844337887234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,8192,0.10795733663770889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,12288,0.10711555348502265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,10240,0.09120000070995754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,7168,0.12231822808583577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,8192,0.05684000253677368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,51200,0.7467661963568794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,7168,0.07611378033955891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,6144,0.1043928861618042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,8192,0.07783022191789415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,7168,0.050423112180497914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,5120,0.08879378106858994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,16384,0.1421413289176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,6144,0.06743644343482123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,7168,0.06903466913435194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,5120,0.059655997488233775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,4096,0.07182933224572076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,6144,0.06070578098297119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,5120,0.05328177743487888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,5120,0.03840533230039809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,4096,0.044941332605150014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,3584,0.06324266725116305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,4096,0.05093244380421109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,4096,0.040116445885764226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,3072,0.0550764434867435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,51200,0.39808977974785703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,3584,0.04489599996142917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,3584,0.028704000843895808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,3072,0.04011999898486667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,3072,0.031705776850382485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,3072,0.03738488753636678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,3584,0.041491554843054876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,2048,0.03674844569630093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,2560,0.03252089023590088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,2560,0.028162668148676556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,1536,0.028656889994939167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,2560,0.033751110235850014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,2048,0.01928622192806668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,65536,0.5502773390875922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,2048,0.029123554627100628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,1024,0.02017688916789161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,2048,0.0302817788388994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,1536,0.01643111142847273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,1536,0.023331556055280898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,768,0.01497244503762987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,1536,0.024650666448805068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,1024,0.014859555496109856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,1024,0.01720533271630605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,512,0.010572444233629437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,1024,0.021754667162895203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,768,0.013601777454217276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,768,0.017532444662517972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,256,0.006510222123728857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,768,0.013788444300492605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,512,0.010836444795131683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,512,0.011525332927703857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,2560,0.045765333705478244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,6144,0.04547911220126682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,512,0.016730666160583496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,256,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,64,0.002579555536309878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,256,0.008214222060309516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,32,0.0028408887899584244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,256,0.013728000223636627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,51200,128,0.009402666654851701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,128,0.007359111474619971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,51200,128,0.010372444159454769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,64,0.008314666648705801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,51200,32,0.008787555827034844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,65536,0.3222737842135959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,65536,0.1766444444656372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,51200,0.24923110008239746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,65536,0.16576266288757324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,51200,0.13116355737050375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,51200,128,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,51200,0.12465777662065293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,12288,0.06567555665969849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,16384,0.04127644499142965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,16384,0.05584177705976698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,12288,0.03400177756945292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,10240,0.05323466658592224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,12288,0.05224800109863281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,12288,0.045032888650894165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,16384,0.08475199672910903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,10240,0.028756442997190688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,51200,0.1373022264904446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,65536,0.17797777387830946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,10240,0.043397333886888295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,7168,0.036680890454186335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,10240,0.040213333235846624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,16384,0.04758666621314155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,8192,0.024354666471481323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,8192,0.04073066512743632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,6144,0.03199555476506551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,8192,0.03419999943839179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,7168,0.032237334383858576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,5120,0.027404444085227117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,7168,0.022184888521830242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,7168,0.03271289004219903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,6144,0.019596444235907663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,6144,0.028200000524520874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,5120,0.017663111289342243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,6144,0.029236445824305218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,5120,0.026194666822751362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,5120,0.025631111529138353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,4096,0.019159111711714003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,4096,0.021332444416152105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,3072,0.01756177842617035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,4096,0.022696000006463792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,3584,0.019344889455371432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,3584,0.017851556340853374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,2560,0.01479200025399526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,3584,0.020582222276263766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,3072,0.013166222307417126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,3072,0.017183999220530193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,2048,0.012548444171746572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,3072,0.019670221540662978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,2560,0.012083555261294046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,2560,0.01517955627706316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,1536,0.00998399986161126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,2560,0.017875555488798354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,2048,0.011008888483047485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,2048,0.013454222016864352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,1024,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,1536,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,4096,0.02200177808602651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,1536,0.01500266624821557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,1024,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,1024,0.013409777647919126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,768,0.008005333443482717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,8192,0.042248000701268516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,768,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,768,0.012102222277058495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,512,0.0042462220622433555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,512,0.007277333074145847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,512,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,256,0.0025128889828920364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,512,0.010682666467295753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,256,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,128,0.0021262222694026097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,256,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,3584,0.0196426659822464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,768,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,64,0.0021342221233579847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,128,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,128,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,16384,32,0.0020311110549502904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,128,0.009007111191749573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,64,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,32,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,65536,0.24550133281283906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,65536,0.1474977731704712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,51200,0.18916355239020455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,1536,0.011301333705584208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,65536,0.1262435515721639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,256,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,16384,1024,0.009319110876984065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,16384,0.0647911098268297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,51200,0.11180267069074844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,51200,0.1006115542517768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,12288,0.050204445918401085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,16384,0.040808889600965716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,16384,0.037440889411502414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,16384,0.04427999920315213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,10240,0.042415112257003784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,12288,0.029663999875386555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,12288,0.043875555197397866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,12288,0.035994665490256414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,10240,0.02568000058333079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,10240,0.035628444618648954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,51200,0.10663111342324151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,10240,0.032636443773905434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,7168,0.030920889642503526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,8192,0.021917333205540974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,8192,0.03576088945070902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,6144,0.02700622214211358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,8192,0.0291884442170461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,7168,0.030957334571414526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,7168,0.01995022263791826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,7168,0.0262755552927653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,5120,0.023620444867346022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,6144,0.017835555805100333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,6144,0.024872889121373493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,4096,0.020358221398459542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,6144,0.02407466702991062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,5120,0.02254755629433526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,5120,0.016173332929611206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,5120,0.021472889516088698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,3584,0.017893332574102614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,65536,0.13771733972761366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,4096,0.014747555057207743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,4096,0.019555555449591745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,3072,0.015846222639083862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,4096,0.019309333629078336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,3584,0.017447999782032438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,8192,0.03558044301138984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,2560,0.013743999931547375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,3584,0.018360000517633226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,3072,0.012468444804350534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,2048,0.01716888944307963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,3072,0.01577333278126187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,2048,0.011937778029176923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,3072,0.01790577835506863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,2560,0.013101333545313941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,1536,0.00793511089351442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,2560,0.014431110686726041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,2048,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,2560,0.016484444340070088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,1024,0.006341333190600078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,2048,0.0125608891248703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,2048,0.015970667203267414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,1536,0.00868888861603207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,1536,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,768,0.005422222117582957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,1536,0.01478666729397244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,1024,0.00833777752187517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,512,0.003547555456558863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,768,0.007320000065697565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,1024,0.012423111332787408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,768,0.0070053330726093715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,256,0.002514666567246119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,768,0.011017777853541903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,512,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,128,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,512,0.011120888921949597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,256,0.006558222075303395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,256,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,64,0.0018880000958840053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,3584,0.013391111459996967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,256,0.00979733301533593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,12288,32,0.0019057778020699818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,128,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,128,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,128,0.009340444372759925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,64,0.005667555663320754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,32,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,65536,0.20486044883728027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,65536,0.14008177651299372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,51200,0.16060533788469103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,65536,0.1279262171851264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,51200,0.10706133312649196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,12288,1024,0.008223999705579545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,51200,0.10112088918685913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,12288,0.042678223715888135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,512,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,16384,0.039537777503331505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,16384,0.03787111242612203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,10240,0.03648533423741659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,16384,0.04045066568586562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,12288,0.028879112667507593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,12288,0.035736001200146146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,51200,0.09411733018027411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,12288,0.03330933385425144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,65536,0.11887999375661214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,8192,0.030453334252039593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,10240,0.030576000610987347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,7168,0.0273644444015291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,10240,0.029872000217437744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,8192,0.020945777495702107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,8192,0.028002666102515325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,6144,0.024051555328898962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,8192,0.027041777968406677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,7168,0.0191928893327713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,7168,0.022915555371178523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,16384,0.057209776507483594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,7168,0.024478221933046978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,6144,0.017320000463061862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,6144,0.02109066645304362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,4096,0.017267554998397827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,5120,0.015656888484954834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,5120,0.018633777896563213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,5120,0.02053777707947625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,4096,0.016361777981122334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,3584,0.015607999430762397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,4096,0.018407111366589863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,10240,0.024259555670950148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,3584,0.015043555034531487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,3584,0.012991999586423239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,3072,0.013906665974193148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,3584,0.018238221605618794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,2560,0.012668444050682915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,3072,0.013062222136391534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,3072,0.013669333524174161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,3072,0.016600888636377122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,5120,0.02086666723092397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,2048,0.010258666343159145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,2560,0.01258933295806249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,2560,0.010393777655230628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,1536,0.007807110746701558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,2560,0.015809777710172866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,2048,0.009364444348547194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,6144,0.022257778379652236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,1024,0.005690666536490123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,2048,0.015092444088723926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,1536,0.008556444611814287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,1536,0.013797333670987023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,768,0.003946666502290302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,1024,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,1024,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,1024,0.011414222419261932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,512,0.002887111157178879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,768,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,768,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,256,0.0022968887868854734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,768,0.011403555671374003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,512,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,512,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,128,0.0021520000365045336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,512,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,256,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,256,0.010748444332016839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,64,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,128,0.006476444502671559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,128,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,10240,32,0.0021297778520319196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,2048,0.010979555547237396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,128,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,65536,0.17122222317589653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,1536,0.009295110901196798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,64,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,10240,32,0.00573244442542394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,4096,0.013762666947311826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,51200,0.12875555621253118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,65536,0.08828888999091254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,16384,0.04351111253102621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,65536,0.07125066386328803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,51200,0.05784000290764702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,12288,0.034373333056767784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,16384,0.02311466634273529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,16384,0.02754844393995073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,16384,0.03830133212937249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,256,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,10240,0.026713778575261433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,12288,0.018553778529167175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,12288,0.027412444353103638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,65536,0.11042222049501206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,8192,0.022839110758569505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,12288,0.03164800008138021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,10240,0.016554666890038382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,10240,0.023689778314696416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,7168,0.019596444235907663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,10240,0.028919110695521038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,8192,0.020920889245139226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,6144,0.01717066599263085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,8192,0.025030222203996446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,7168,0.013491555220550962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,7168,0.018241778016090393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,5120,0.014771555860837301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,7168,0.023044443792766992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,6144,0.012706666356987424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,51200,0.06903733147515191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,6144,0.018014222383499146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,4096,0.01313866674900055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,6144,0.021336888273557026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,5120,0.011731555892361535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,5120,0.014466666513019137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,3584,0.010944888823562197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,5120,0.019909333851602342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,4096,0.013062222136391534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,4096,0.010703110860453712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,3072,0.009932444327407414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,4096,0.017941332525677152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,51200,0.0891688863436381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,3584,0.009305777649084726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,3584,0.012079999678664737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,2560,0.008504889077610439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,3584,0.01759733259677887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,3072,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,3072,0.011384000380833944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,2048,0.007352888584136963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,8192,0.01477777792347802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,3072,0.0162462227874332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,2560,0.00831733312871721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,2560,0.010260444548394945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,1536,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,2560,0.015495111544926962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,2048,0.007602666815121968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,1024,0.005092444519201915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,2048,0.009367999931176504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,1536,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,768,0.003548444559176763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,1536,0.007686222592989604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,1536,0.012758221891191272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,1024,0.006634666687912411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,512,0.0025724443710512584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,1024,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,1024,0.011026666396194033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,768,0.006299555715587404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,768,0.006152888966931238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,256,0.0022355556074115965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,768,0.011072888970375061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,512,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,128,0.002114666625857353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,512,0.010334221853150262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,256,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,256,0.0053013331360287136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,64,0.0019084444890419643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,256,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,8192,32,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,128,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,128,0.008264000217119852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,65536,0.14753333727518717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,64,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,32,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,51200,0.11617333359188503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,2048,0.014244443840450711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,65536,0.06581155459086101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,51200,0.0626773304409451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,16384,0.03974577784538269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,8192,512,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,65536,0.10986755291620891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,16384,0.021634666456116572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,16384,0.02498044404718611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,12288,0.03155288762516446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,128,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,51200,0.08902311325073242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,10240,0.026452443665928308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,16384,0.037754668129815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,12288,0.017664889494578045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,8192,0.02253599961598714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,12288,0.026055110825432673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,10240,0.015800888339678448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,12288,0.03118577930662367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,7168,0.017847110827763874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,10240,0.028483556376563177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,65536,0.08249066935645209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,8192,0.014125333891974555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,8192,0.02034133341577318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,6144,0.015489778584904142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,8192,0.02453777856296963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,7168,0.01291911138428582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,7168,0.017539555827776592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,5120,0.013464888764752282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,7168,0.023002665903833177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,6144,0.01737511157989502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,4096,0.011703111231327057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,6144,0.021425777011447485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,5120,0.010886222124099731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,5120,0.014120888378885059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,3584,0.010070222119490305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,5120,0.019187556372748483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,4096,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,4096,0.01274577776590983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,3072,0.009189333352777693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,3584,0.008997333546479544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,4096,0.018201778332392376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,3584,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,2560,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,3584,0.01683466633160909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,3072,0.008616888688670265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,10240,0.02234488891230689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,3072,0.016755556066830952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,2048,0.006551110910044775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,2560,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,2560,0.01568711135122511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,6144,0.012079111403889127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,2048,0.008954666554927826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,1536,0.005650666852792104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,51200,0.05347822109858195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,2048,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,2048,0.014412444498803882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,1024,0.0038524443904558816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,1536,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,1536,0.00756888919406467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,768,0.003208000005947219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,1024,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,1536,0.013470222552617392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,1024,0.011711111499203576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,512,0.002552888873550627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,768,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,768,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,768,0.010551999840471478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,256,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,512,0.005960888746711943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,512,0.005648889061477449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,2560,0.007376000285148621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,512,0.011214222345087262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,128,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,256,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,256,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,64,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,256,0.010485333700974783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,7168,32,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,128,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,128,0.004965333475006951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,128,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,65536,0.12097688515981038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,64,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,32,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,51200,0.09540888998243545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,65536,0.07407022184795804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,1024,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,16384,0.03377866744995117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,65536,0.06323733594682482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,51200,0.056419557995266385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,51200,0.051368888881471425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,12288,0.026936888694763184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,7168,3072,0.011344888971911537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,16384,0.023503111468421087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,16384,0.02086044516828325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,10240,0.022974222898483276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,65536,0.10964177714453803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,16384,0.03759288787841797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,51200,0.08856978019078572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,12288,0.01720888912677765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,12288,0.023688889212078516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,8192,0.01972888906796773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,7168,0.01754577789041731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,12288,0.030829333596759375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,10240,0.015238222148683337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,10240,0.020671111014154222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,8192,0.013412444127930535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,6144,0.015477332803938123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,10240,0.02780710988574558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,8192,0.018412444326612685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,8192,0.024015110399987962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,5120,0.013740444348918067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,7168,0.016173332929611206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,6144,0.011751111182901593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,7168,0.022320000661744013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,6144,0.015113777584499784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,5120,0.0107004443804423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,5120,0.013795555465751223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,3584,0.009334222310119206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,4096,0.008771555291281806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,4096,0.012381333443853589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,4096,0.017479999197853934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,3072,0.0079297779334916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,3584,0.008480889101823172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,3584,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,3584,0.01685599982738495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,2560,0.007228444847795699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,3072,0.010503110786279043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,3072,0.008003555238246918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,3072,0.015867556134859722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,2048,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,2560,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,2560,0.009672000176376766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,7168,0.012549333274364471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,2560,0.0151555554734336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,1536,0.005323555734422471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,2048,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,4096,0.01036177741156684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,1024,0.0034497777620951333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,6144,0.02106222179200914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,1536,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,1536,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,768,0.0028817777832349143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,1536,0.012375999655988483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,5120,0.01912088857756721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,1024,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,1024,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,512,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,1024,0.011487111449241638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,768,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,768,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,768,0.010486221975750394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,512,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,512,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,128,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,512,0.010684444672531553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,256,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,256,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,64,0.0018719999740521114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,256,0.009740444521109263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,128,0.005636444522274866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,6144,32,0.001889777680238088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,128,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,2048,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,65536,0.10571733448240493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,64,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,2048,0.013502221968438892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,6144,32,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,51200,0.08309600088331434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,65536,0.06638933552636041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,16384,0.030271109607484605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,51200,0.05089955528577169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,51200,0.049244443575541176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,16384,0.02121688922246297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,65536,0.10954578055275811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,16384,0.020340444313155282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,12288,0.024358222881952923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,16384,0.03716266817516751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,10240,0.020599111914634705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,12288,0.016794666647911072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,12288,0.025246222813924152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,12288,0.030198223061031763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,8192,0.018057778477668762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,10240,0.014831110835075378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,10240,0.01945955554644267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,10240,0.026862222287389968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,7168,0.015793777174419828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,8192,0.013122667041089801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,8192,0.016984888248973422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,8192,0.02370400064521366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,65536,0.06112000015046862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,6144,0.014154665999942355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,7168,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,7168,0.01589777734544542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,7168,0.02236622240808275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,5120,0.012244444754388599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,6144,0.011066666907734342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,6144,0.01475822263293796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,51200,0.0886675582991706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,4096,0.009666666388511658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,6144,0.020273778173658583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,5120,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,3584,0.008612444831265343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,5120,0.012772444221708508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,4096,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,5120,0.018872888551818002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,4096,0.011726222104496427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,4096,0.017200888858901132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,3584,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,3584,0.010792888700962067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,2560,0.006897777732875612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,128,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,3584,0.01720355616675483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,2048,0.005738666488064661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,3072,0.009649777577983009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,3072,0.015834665960735746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,2560,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,1536,0.0047866664826869965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,2560,0.008649778034951951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,2560,0.014089778065681458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,2048,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,1024,0.003253333270549774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,2048,0.012720888687504662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,1536,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,1536,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,768,0.0030257778449190986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,1536,0.012372444073359171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,1024,0.0061919999619325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,1024,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,512,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,1024,0.011022222538789114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,768,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,768,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,256,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,768,0.011128000087208219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,512,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,3072,0.007925333248244392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,128,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,512,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,256,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,256,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,64,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,3072,0.0076248885856734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,256,0.01034311122364468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,128,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,5120,32,0.0018159999615616268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,128,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,128,0.008015111088752747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,2048,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,64,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,32,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,51200,0.06541955471038818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,65536,0.059191114372677274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,65536,0.058825777636633984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,16384,0.02477333280775282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,51200,0.04645244280497233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,51200,0.04757600029309591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,65536,0.1095199982325236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,16384,0.01959822244114346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,51200,0.0885475542810228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,5120,512,0.0052888890107472735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,16384,0.036481777826944985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,10240,0.016706667012638517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,12288,0.016215110818545025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,12288,0.0203475554784139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,12288,0.029539555311203003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,8192,0.014125333891974555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,10240,0.01443377799457974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,65536,0.08937244282828437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,10240,0.019208888212839764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,7168,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,10240,0.02625511089960734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,8192,0.012836444709036084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,8192,0.01680177781316969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,6144,0.011731555892361535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,8192,0.02349688940578037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,7168,0.01551644504070282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,7168,0.011693333586057028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,5120,0.01051288925939136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,6144,0.01038755559258991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,6144,0.014138667119873894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,12288,0.019215111931165058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,16384,0.0192248887485928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,4096,0.009318222602208456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,5120,0.012088889049159156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,5120,0.008716444174448649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,3584,0.010809777511490716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,5120,0.01885511146651374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,4096,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,4096,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,3072,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,4096,0.016883555385801528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,3584,0.010058666268984476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,3584,0.01623644431432088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,2560,0.008396444221337637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,3072,0.0074782222509384155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,3072,0.008616888688670265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,2048,0.0069982219073507525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,3072,0.014929778046078153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,2560,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,2560,0.00812355594502555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,1536,0.005664000080691443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,7168,0.021909332937664453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,2048,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,2048,0.007239110767841339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,1024,0.004201777693298128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,2048,0.012658666405412885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,1536,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,1536,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,6144,0.020246222615242004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,768,0.0038311113086011675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,1536,0.011738667057620155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,1024,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,1024,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,512,0.002943111169669363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,1024,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,768,0.006021333237489064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,768,0.005640000104904175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,256,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,512,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,768,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,3584,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,512,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,128,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,512,0.010727110836240979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,256,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,128,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,32,0.0018604445374674266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,256,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,128,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,128,0.008665777742862701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,65536,0.07653244336446126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,64,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,2560,0.01350844403107961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,32,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,65536,0.057646221584743924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,65536,0.05783200263977051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,16384,0.022721777359644573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,51200,0.044778668218188815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,51200,0.04662755462858412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,51200,0.08811910947163899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,16384,0.018146667215559218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,12288,0.017525333497259352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,16384,0.01921866668595208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,16384,0.035904887649748064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,4096,64,0.001820444439848264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,10240,0.015475556254386902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,4096,256,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,12288,0.016100444727473788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,12288,0.029574222034878198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,8192,0.012813333008024426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,10240,0.014176888598336114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,10240,0.01793244481086731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,51200,0.05972533093558418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,10240,0.02648711038960351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,7168,0.012040889097584618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,8192,0.015829333000712924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,8192,0.012358222570684222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,6144,0.011009777585665384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,8192,0.02332177758216858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,7168,0.011133333047231039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,7168,0.013358222113715278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,65536,0.10939644442664252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,5120,0.009712888962692684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,6144,0.010299555957317352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,6144,0.013056889176368713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,4096,0.008763555851247575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,6144,0.020242666204770405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,5120,0.009326222042242685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,5120,0.01202488856183158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,3584,0.010240889257854885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,5120,0.0185253338681327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,12288,0.021478222476111516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,4096,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,4096,0.010249777800507015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,3072,0.009294222626421187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,3584,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,3584,0.00868444475862715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,2560,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,3072,0.00832799987660514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,3072,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,3584,0.015799111790127225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,3072,0.014440000057220459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,2048,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,2560,0.007116444408893585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,2560,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,1536,0.0053191110491752625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,2560,0.013733333183659447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,2048,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,2048,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,1024,0.0041911109454101985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,2048,0.0127368892232577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,7168,0.02164977788925171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,1536,0.00629066675901413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,1536,0.006666666517655055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,768,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,1536,0.012309333516491784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,1024,0.006157333238257303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,512,0.0028684443483750024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,1024,0.006190222170617845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,1024,0.011027555498811932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,768,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,768,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,768,0.010457778142558204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,512,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,128,0.0018613334331247543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,4096,0.016504888733228047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,256,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,256,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,128,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,128,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,128,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,65536,0.06161777840720283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,64,0.005003555367390315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3584,32,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,51200,0.048900445302327476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,65536,0.05693778064515856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,65536,0.05440444416469998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,65536,0.10923200183444554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,51200,0.04279555545912849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,51200,0.045909331904517285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,16384,0.018872888551818002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,256,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,512,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,51200,0.08816711107889812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,16384,0.018927110566033255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,512,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,16384,0.01756088932355245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,10240,0.013606222139464485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,16384,0.03567377726236979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,12288,0.01568266749382019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,8192,0.011737777955002256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3584,32,0.0018275556051068837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,12288,0.029263112280103896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,10240,0.013783999615245394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,7168,0.010969777901967367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,10240,0.026572444372706946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,8192,0.011755555868148804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,8192,0.014999111493428549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,6144,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,8192,0.02368533280160692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,7168,0.010382222632567087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,7168,0.01315199997689989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,7168,0.021929777330822412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,5120,0.008795555267069075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,6144,0.011720889144473605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,6144,0.020230222079488967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,12288,0.015830222103330824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,5120,0.010677333507272931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,4096,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,5120,0.008952889177534316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,5120,0.019009777241282992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,3584,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,4096,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,12288,0.01696266730626424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,4096,0.009223110973834991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,3072,0.008596444295512306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,4096,0.016160888804329764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,3584,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,3584,0.00797155582242542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,2560,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,3584,0.01514755520555708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,3072,0.007669332954618666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,3072,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,2048,0.006224888894293044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,3072,0.014142221874660917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,2560,0.007342221836249034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,2560,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,1536,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,6144,0.009468444519572789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,2560,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,2048,0.006597333484225803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,2048,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,1024,0.0041866666740841335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,2048,0.0127368892232577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,1536,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,1536,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,768,0.0038693332009845306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,1536,0.011906666888131036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,1024,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,1024,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,512,0.0031662223239739737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,1024,0.01142133358452055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,256,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,768,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,768,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,512,0.0058871110280354815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,128,0.002191999927163124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,512,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,10240,0.017158221867349412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,256,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,64,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,256,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,128,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,3072,32,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,256,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,128,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,65536,0.05553333295716179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,128,0.008003555238246918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,64,0.004948444250557157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,51200,0.04288888971010844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,3072,32,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,65536,0.05615288681454129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,65536,0.05225155419773526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,16384,0.017337777548366122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,51200,0.04516177707248264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,51200,0.041442665788862444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,65536,0.10911999808417426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,12288,0.014188443620999655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,51200,0.0880791081322564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,16384,0.01719199948840671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,16384,0.018644443816608854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,10240,0.012063110868136087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,16384,0.035387555758158364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,12288,0.015119999647140503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,12288,0.01646222174167633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,8192,0.010518222219414182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,10240,0.013336000343163809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,768,0.010386666489972008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,10240,0.015229332778188916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,7168,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,512,0.010971555279360877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,10240,0.026320000489552815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,8192,0.011448889142937131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,6144,0.008658666577604082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,8192,0.01312800000111262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,7168,0.01038577738735411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,8192,0.023165333602163527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,7168,0.012347555822796293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,5120,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,7168,0.021389333738221064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,6144,0.010890666809346942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,4096,0.0074782222509384155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,6144,0.020079111059506733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,5120,0.009665778113736046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,3584,0.006903999795516332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,5120,0.018010665973027546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,4096,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,4096,0.00793333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,3072,0.007520889242490132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,4096,0.016176000237464905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,3584,0.007742221984598372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,3584,0.007272889216740926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,3584,0.01484266585773892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,2560,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,3072,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,3072,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,12288,0.029493331909179688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,2048,0.005883555445406172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,2560,0.007241778075695038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,2560,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,2560,0.013479111095269522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,1536,0.005190222213665645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,2048,0.00664533343580034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,6144,0.009378666679064432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,2048,0.012431110772821637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,1024,0.004101333518822988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,5120,0.009022222624884712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,1536,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,1536,0.005978666659858491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,768,0.0035297779573334586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,1536,0.011754666765530905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,1024,0.006224888894293044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,1024,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,512,0.0028604444944196274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,1024,0.011445333560307821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,768,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,768,0.005283555636803309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,256,0.002493333278430833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,768,0.01034577770365609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,512,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,512,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,128,0.0021422221842739317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,3072,0.014140443669425117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,256,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,256,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,256,0.009826666778988307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,128,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,128,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,2048,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,65536,0.047988444566726685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,128,0.008035555481910706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,32,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,51200,0.03700977894994948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,65536,0.10922666390736897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,51200,0.04008888867166307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,51200,0.044756445619795054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,16384,0.014099554883109199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,51200,0.08824177583058675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,512,0.01073155552148819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,16384,0.018390221728218924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,16384,0.01664355562792884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,16384,0.035232888327704534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2560,32,0.001806222316291597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,10240,0.01053955571519004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,12288,0.014871110518773397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,12288,0.014664888381958008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,12288,0.029143999020258587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,8192,0.009280000295903947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,10240,0.01204533295498954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,10240,0.014627556006113688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,65536,0.05009777678383721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,7168,0.008994667066468133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,10240,0.026312000221676294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,8192,0.0107351111041175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,8192,0.012430222498046027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,6144,0.00777599960565567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,7168,0.00999644481473499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,7168,0.01015466699997584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,7168,0.021734222769737244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,5120,0.007133333219422235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,6144,0.009700444837411245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,6144,0.009750222166379293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,12288,0.012063999970753988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,6144,0.01920977731545766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,5120,0.00905777762333552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,5120,0.00833777752187517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,5120,0.017720000611411203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,3584,0.008107555409272512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,4096,0.007627555893527136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,4096,0.008287111090289222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,4096,0.01572711103492313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,3072,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,65536,0.05526577764087253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,3584,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,3584,0.014817777607176038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,2560,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,3072,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,3072,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,8192,0.022568888134426538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,3072,0.013997332917319404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,2048,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,2560,0.007017778025733099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,2560,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,1536,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,2560,0.013961777091026306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,2048,0.006912888752089606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,2048,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,4096,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,1024,0.0038968887594011095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,2048,0.012727999852763282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,1536,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,1536,0.005873777800136142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,768,0.003543111185232798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,1536,0.012432888978057437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2560,64,0.004942222187916438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,1024,0.006216889040337668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,1024,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,512,0.0028906667398081887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,1024,0.011023999916182624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,768,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,3584,0.00795199970404307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,512,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,768,0.011056000159846412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,512,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,128,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,512,0.01019022199842665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,256,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,256,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,64,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,256,0.010017777482668558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,128,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,128,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,32,0.0019786666250891155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,65536,0.03432088759210374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,128,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,64,0.004962666581074397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,51200,0.027258666025267705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,32,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,65536,0.05464622378349304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,65536,0.048687110344568886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,16384,0.012056888805495368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,65536,0.10902933279673259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,51200,0.04428177740838793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,51200,0.08765333228641087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,16384,0.01692800058258904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,2048,256,0.002471111093958219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,16384,0.035024000538720026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,12288,0.014289778139856128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,12288,0.013811555173661975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,10240,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,12288,0.028961777687072754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,10240,0.012072888513406118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,10240,0.01240799989965227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,10240,0.025976000560654536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,8192,0.010679999987284342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,51200,0.038698666625552706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,8192,0.022293332550260756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,7168,0.00795288880666097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,8192,0.010687111152542962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,2048,768,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,12288,0.009710222482681274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,7168,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,7168,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,5120,0.006220444622966979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,7168,0.020614221692085266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,6144,0.009374221993817223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,6144,0.008804444637563493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,4096,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,6144,0.019034667147530448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,5120,0.008625778059164682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,5120,0.008257778154479133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,3584,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,5120,0.017509332961506315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,8192,0.0075795559419526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,4096,0.007685333490371704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,4096,0.008003555238246918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,3072,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,4096,0.016148444679048326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,3584,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,3584,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,2560,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,3584,0.01517777807182736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,3072,0.007631111476156447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,2048,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,3072,0.014470222923490735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,2560,0.007048000064161088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,2560,0.006520000182920032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,16384,0.016167110866970487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,1536,0.004986666556861666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,2560,0.013444444371594323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,6144,0.0074444446298811175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,1024,0.003548444559176763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,2048,0.006687110910813014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,2048,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,1536,0.006131555471155379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,2048,0.01310755560795466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,768,0.0031146665828095544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,1536,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,1536,0.012028444144460889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,1024,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,512,0.002528000002106031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,1024,0.011386666860845355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,768,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,256,0.0021946666141351065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,768,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,512,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,768,0.010348444183667501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,128,0.0021377777059872947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,512,0.010203555226325989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,256,0.00489688871635331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,256,0.010397333237859938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,3072,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,128,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1536,32,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,128,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,65536,0.02407022151682112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,128,0.008388444781303406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,64,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,32,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,51200,0.01958933307064904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,65536,0.05449155635303921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,65536,0.04695111182000902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,16384,0.013346666263209449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,65536,0.10868799686431885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,1024,0.005315555466545953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,51200,0.04383733206325107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,12288,0.010695111420419482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,51200,0.03750577900144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,16384,0.015159999330838522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,51200,0.08697777986526489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,10240,0.00998488896422916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1536,512,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,16384,0.034910221894582115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,12288,0.012446222205956778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,12288,0.01310400002532535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,8192,0.00960000024901496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,12288,0.0281644430425432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,10240,0.012074666718641916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,10240,0.01127822200457255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,7168,0.008650666309727563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,8192,0.010478222535716163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,8192,0.009752000371615091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,10240,0.025516443782382544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,6144,0.007140444384680853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,8192,0.022442667020691767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,7168,0.009735999835862054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,5120,0.006525333142942852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,7168,0.020613332589467365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,6144,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,6144,0.008763555851247575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,4096,0.005877333382765452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,6144,0.018518222702874076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,5120,0.0081404447555542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,5120,0.017508443858888414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,4096,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,3584,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,4096,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,16384,0.01606044504377577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,3072,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,4096,0.015764445066452026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,3584,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,3584,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,2560,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,3584,0.015496888094478183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,3072,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,3072,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,2048,0.00609688874748018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,3072,0.013791110780504016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,2560,0.007280888656775157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,2560,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,1536,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,2560,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,7168,0.009333333207501305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,2048,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,2048,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,1024,0.0037466668420367767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,2048,0.012537777423858643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,768,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,1536,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,1024,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,1536,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,512,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,1536,0.012087110843923358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,1024,0.010724444356229572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,768,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,256,0.002432000098956956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,768,0.010406222608354356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,512,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,512,0.004961777892377642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,128,0.002163555473089218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,512,0.011054221954610614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,256,0.0052977779673205475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,256,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,256,0.009249777429633671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,128,0.0052879999081293745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,128,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,1024,32,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,65536,0.019208000765906442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,128,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,64,0.004856889032655292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,51200,0.01627111103799608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,32,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,65536,0.05381422241528829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,65536,0.046012444628609546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,16384,0.013447999954223633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,5120,0.008709333009190029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,51200,0.04348444607522753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,51200,0.0369484457704756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,1024,1024,0.00545155546731419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,12288,0.010049777726332346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,768,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,51200,0.08656799793243408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,16384,0.015788444214397006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,10240,0.008515555825498369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,16384,0.015106666419241162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,12288,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,16384,0.034584889809290566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,12288,0.012223111258612739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,8192,0.007446222007274628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,12288,0.028869334194395278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,10240,0.011006222003036074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,7168,0.0069368887278768755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,10240,0.025693333811230127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,8192,0.010692444112565784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,8192,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,6144,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,8192,0.021990221407678392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,7168,0.010057777994208867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,7168,0.009339555270142024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,5120,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,7168,0.020566221740510728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,6144,0.009385777844323052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,6144,0.008646222452322641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,4096,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,6144,0.018887110882335238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,65536,0.1083128915892707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,5120,0.008972444468074376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,5120,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,3584,0.005624889085690181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,5120,0.01755555636352963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,4096,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,4096,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,3072,0.005492444253630108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,4096,0.016191111670600045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,3584,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,3584,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,2560,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,3584,0.01516355574131012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,3072,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,3072,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,2048,0.0042133331298828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,10240,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,3072,0.014017777310477363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,1536,0.004502222355869082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,2560,0.007253333098358578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,2560,0.006446222050322427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,2048,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,2560,0.013741333451535968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,1024,0.0035217776894569397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,2048,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,1536,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,2048,0.013095999757448832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,1536,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,1024,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,1536,0.012039999994966718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,1024,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,512,0.0027635554886526535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,1024,0.010712888505723743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,768,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,768,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,256,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,768,0.01036177741156684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,512,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,512,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,128,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,512,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,256,0.005388444496525659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,64,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,32,0.0018088887963030075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,256,0.009644444617960189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,128,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,65536,0.027038221557935078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,128,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,64,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,32,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,51200,0.022301332818137273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,65536,0.051635556750827365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,65536,0.1076471143298679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,16384,0.0090515555606948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,51200,0.038395557138654925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,51200,0.03642311029964023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,51200,0.08630310826831394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,12288,0.007520000139872233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,16384,0.015815999772813585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,16384,0.014449778530332776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,16384,0.034280889564090304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,10240,0.006786666810512543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,12288,0.013064888616402945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,12288,0.012196444802814059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,256,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,12288,0.027738667196697656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,8192,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,10240,0.011754666765530905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,768,128,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,10240,0.01105955574247572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,7168,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,10240,0.02472355630662706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,8192,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,8192,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,6144,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,8192,0.021967111362351313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,65536,0.04550844430923462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,7168,0.009700444837411245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,7168,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,5120,0.005304000029961268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,7168,0.01995644470055898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,768,768,0.002982222164670626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,6144,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,4096,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,6144,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,5120,0.00832799987660514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,6144,0.01885777711868286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,3584,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,5120,0.016866667403115165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,4096,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,3072,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,4096,0.015448888142903646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,3584,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,3584,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,2560,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,3584,0.0151591118839052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,3072,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,3072,0.006649777707126405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,2048,0.004592000196377437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,3072,0.013809777796268463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,2560,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,2560,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,1536,0.003524444583389494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,2560,0.013735999663670858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,2048,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,2048,0.006139555739031897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,1024,0.0028764444092909494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,2048,0.01237511138121287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,1536,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,1536,0.012090666426552666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,768,0.003076444483465619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,1024,0.006265777680608962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,1024,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,1024,0.01072355525361167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,4096,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,512,0.002480000050531493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,768,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,768,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,256,0.0021724444296624926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,768,0.011076444553004371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,512,0.005635555419656966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,512,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,128,0.0020879999630981022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,512,0.010028444230556488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,256,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,256,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,64,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,256,0.010089777410030365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,512,32,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,128,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,65536,0.016139555308553908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,128,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,64,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,1536,0.006032889087994893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,512,32,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,51200,0.014298665854665967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,65536,0.04689244429270426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,65536,0.04220889012018839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,16384,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,5120,0.008615111311276754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,65536,0.10728355248769124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,51200,0.03735377722316318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,12288,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,51200,0.03404622276624044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,16384,0.01579466627703773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,51200,0.08630755874845718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,16384,0.014365333649847241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,10240,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,12288,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,16384,0.03433600068092346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,8192,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,10240,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,12288,0.027687110834651526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,7168,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,10240,0.011056000159846412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,8192,0.010384000009960597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,8192,0.009766221874290043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,128,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,8192,0.021313778228229944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,7168,0.009653333160612319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,7168,0.00902133352226681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,5120,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,7168,0.020629333125220407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,6144,0.009312888814343346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,6144,0.008844444321261512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,4096,0.004547555413511065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,6144,0.018231110440360177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,5120,0.008559111091825698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,5120,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,3584,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,5120,0.01754755609565311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,4096,0.007674666742483775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,4096,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,4096,0.015513777732849121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,3072,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,3584,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,3584,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,12288,0.012104888757069906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,2560,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,3584,0.014761777387724983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,3072,0.007345777418878343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,3072,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,10240,0.024745777249336243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,2048,0.004569777597983678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,2560,0.00721866637468338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,6144,0.005195555587609609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,3072,0.01444533301724328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,1536,0.003546666767862108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,2560,0.01313777764638265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,2048,0.006783999916579988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,2048,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,1024,0.0028524444335036804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,1536,0.006015111174848344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,2048,0.012413333687517377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,768,0.0028106667515304354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,1536,0.012039110892348819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,1024,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,1024,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,512,0.002471111093958219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,1024,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,768,0.005693333430422678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,768,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,256,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,768,0.01072177787621816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,512,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,512,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,128,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,512,0.011044444309340583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,256,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,256,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,64,0.0018079999006456798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,256,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,128,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,256,32,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,65536,0.011423110961914062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,128,0.00462311092350218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,2560,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,128,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,51200,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,32,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,65536,0.04244977898067898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,1536,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,16384,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,65536,0.10641155640284221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,51200,0.03419377737575107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,51200,0.037309332026375666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,12288,0.005191111316283544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,16384,0.015835555063353646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,51200,0.08524622519810994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,16384,0.01444355481200748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,10240,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,16384,0.033907555871539645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,12288,0.013049778011110095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,12288,0.01221244451072481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,8192,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,10240,0.011699555648697747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,12288,0.0281057788266076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,10240,0.011077333655622272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,7168,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,10240,0.024671110841963027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,8192,0.021264889174037512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,6144,0.004855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,256,64,0.004566222015354368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,7168,0.020599111914634705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,65536,0.04589600033230252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,5120,0.004552888787455029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,6144,0.009360888765917884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,6144,0.0084915558497111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,6144,0.01829155617290073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,4096,0.0042160000238153665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,5120,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,5120,0.017234666479958426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,4096,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,3584,0.004638222356637319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,4096,0.008192000289758047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,8192,0.01038577738735411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,3584,0.006663110935025745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,3072,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,3584,0.015192000402344598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,3584,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,7168,0.009666666388511658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,7168,0.009340444372759925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,2560,0.004544888933499654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,3072,0.007251555720965068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,3072,0.006604444649484422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,2048,0.004219555606444676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,2560,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,3072,0.014454222387737699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,2560,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,1536,0.0035111109415690103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,5120,0.00829777783817715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,2048,0.006664888726340399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,2048,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,2048,0.013077333569526672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,1024,0.0028613333900769553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,1536,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,1536,0.005656888915432825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,4096,0.01514933341079288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,1536,0.012200888660218982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,768,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,1024,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,1024,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,8192,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,512,0.0024871110088295406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,1024,0.01036800030204985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,768,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,768,0.010404444403118556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,512,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,512,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,512,0.01070577816830741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,128,0.001810666587617662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,256,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,256,0.004616888860861461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,256,0.00871733327706655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,64,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,128,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,32,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,128,0.004952888935804367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,65536,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,128,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,51200,0.006426666759782367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,64,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,16384,0.0052915554907586836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,128,32,0.004616888860861461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,12288,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,65536,0.042027556233935885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,10240,0.004583111239804162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,51200,0.03392444385422601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,8192,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,768,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,16384,0.014113777213626437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,7168,0.004562666846646203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,12288,0.011907555162906647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,6144,0.004554666578769684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,5120,0.004224888980388641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,10240,0.011001778145631155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,4096,0.0042462220622433555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,8192,0.00964355551534229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,3584,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,7168,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,3072,0.004233777936961916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,6144,0.008344888687133789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,2560,0.00423200014564726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,5120,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,2048,0.0038355555799272326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,4096,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,1536,0.0035111109415690103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,3584,0.007241778075695038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,1024,0.0028204443968004654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,3072,0.006454222318198945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,768,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,512,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,2048,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,256,0.0017875555074877208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,1024,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,128,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,512,0.004984000076850255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,64,0.001632888904876179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,256,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,64,32,0.0017884444031450485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,128,0.00489688871635331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,65536,0.0069013333155049225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,128,256,0.002128888956374592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,64,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,51200,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,16384,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,32,0.004512000001139111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,12288,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,65536,0.042134222057130605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,10240,0.004551110996140374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,51200,0.03356000118785434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,2560,0.013079110946920184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,16384,0.014466666513019137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,7168,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,12288,0.011739555332395764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,6144,0.004509333521127701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,2560,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,5120,0.004003555410438114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,8192,0.009379555781682333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,7168,0.009116444322797988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,4096,0.0038684445122877755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,3584,0.0042444442709287005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,768,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,6144,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,3072,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,2560,0.004209777961174647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,5120,0.007700444095664554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,2048,0.00408266650305854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,4096,0.007486222518814935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,1536,0.0034871109657817413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,3072,0.0064293332397937775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,2560,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,1024,0.002817777709828483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,2048,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,768,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,1536,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,512,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,1024,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,8192,0.004168888760937585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,256,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,768,0.004954666727119022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,128,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,64,0.001503999965886275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,512,0.004849777867396672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,10240,0.010767999622556897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,8,32,32,0.0017866666118303935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,256,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,128,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,64,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,32,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,32,3584,0.006993778049945831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,16384,0.3369679980807834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,16384,0.23140533765157065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,51200,0.4868275324503581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,12288,0.25663910971747506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,51200,1.012137730916341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,10240,0.20131288634406197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,12288,0.10407288869222005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,12288,0.15048889319101968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,8,64,1536,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,8192,0.153111113442315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,51200,0.40324089262220597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,10240,0.08732266558541192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,10240,0.12661688857608372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,8192,0.07190666596094768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,7168,0.13228710492451987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,6144,0.11389688650767009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,7168,0.06273866362041898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,5120,0.09583377838134766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,7168,0.10192000203662449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,6144,0.05510400070084465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,6144,0.0872337751918369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,4096,0.0809440016746521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,16384,0.1366222169664171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,3584,0.06817422310511272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,5120,0.0468426677915785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,5120,0.07327822181913587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,3072,0.05886844131681654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,4096,0.047820445564058095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,4096,0.0664586689737108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,2560,0.049954666031731494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,3584,0.041264888313081526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,2048,0.040976001156700984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,3584,0.057429333527882896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,3072,0.04650133185916477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,1536,0.03178222311867608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,2560,0.027388445205158655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,2560,0.039057777987586126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,1024,0.022670222653283015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,2048,0.037179556157853864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,768,0.017867555220921833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,1536,0.027560888065232172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,1024,0.017838221457269456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,8192,0.11752000119951035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,512,0.01276088919904497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,1024,0.02067466742462582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,256,0.0075288886825243635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,512,0.012950222525331708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,768,0.015152888165579902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,128,0.005022222383154763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,512,0.013100444442696042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,256,0.010384889112578498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,64,0.0041680000722408295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,256,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,128,0.01036266651418474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,65536,32,0.004182222402758068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,128,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,64,0.008966222405433655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,3072,0.031368888086742826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,32,0.009388444324334463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,2048,0.02281422250800663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,65536,1536,0.01886577738655938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,65536,0.5125653478834364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,16384,0.2494328816731771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,51200,0.38867643144395614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,65536,768,0.017271111408869427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,65536,0.9716257519192166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,51200,0.7472044626871744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,12288,0.19131022029452852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,16384,0.16922310988108316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,12288,0.1191448900434706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,16384,0.1111253367529975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,10240,0.15958133008744982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,8192,0.1385644409391615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,10240,0.07134489218393962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,12288,0.08190666966968112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,51200,0.3293813334570991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,7168,0.12156000402238633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,65536,0.4347688886854384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,6144,0.10489155848821004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,8192,0.0587751136885749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,7168,0.050556444459491305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,7168,0.079184889793396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,5120,0.08848622110154893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,6144,0.07049510876337688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,4096,0.07147644625769721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,5120,0.038776000340779625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,5120,0.060198220941755504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,3584,0.06286133660210504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,4096,0.041315555572509766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,3072,0.054449776808420815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,3584,0.029143999020258587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,3584,0.044758223825030856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,2560,0.04567289021280077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,10240,0.10258133543862237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,3072,0.02608088817861345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,3072,0.03844177722930908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,2048,0.0366559988922543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,8192,0.09614488813612197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,2560,0.02232444451914893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,2560,0.032229334115982056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,1536,0.02875022093454997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,1024,0.020239111449983384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,2048,0.01960266629854838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,2048,0.030321776866912842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,1536,0.016493333710564505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,768,0.01494488947921329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,1024,0.015453333655993143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,6144,0.04566133353445265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,512,0.010468444062603844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,1024,0.01683022247420417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,768,0.013595555391576556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,256,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,1536,0.02334755493534936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,768,0.015054222610261707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,256,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,512,0.011487111449241638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,4096,0.055143998728858105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,64,0.002532444480392668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,32,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,128,0.00941777808798684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,128,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,64,0.008294222255547842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,32,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,65536,0.31939999262491864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,65536,0.16857155164082846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,51200,0.24892799059549967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,16384,0.0850142240524292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,51200,0.1306017769707574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,51200,128,0.0031582222630580268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,51200,0.11776178412967259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,12288,0.06532977686987983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,16384,0.04314844475852119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,51200,256,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,51200,512,0.011009777585665384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,16384,0.061483555369906955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,8192,0.04254577888382805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,12288,0.050161778926849365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,12288,0.03328533305062188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,10240,0.028913779391182795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,7168,0.03589866558710734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,10240,0.04285600119166904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,6144,0.03149866726663377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,8192,0.024134222004148696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,8192,0.034671111239327326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,7168,0.021923555268181696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,5120,0.027368000812000696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,7168,0.03136355678240458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,4096,0.022828444838523865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,6144,0.01982933282852173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,6144,0.028869334194395278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,3584,0.01955911186006334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,5120,0.02403644389576382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,4096,0.015488889482286243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,3072,0.01737422247727712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,4096,0.021839999490314062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,65536,0.15340267287360296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,3584,0.018834667073355783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,3072,0.013354666531085968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,3072,0.016414221790101793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,2048,0.01239999963177575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,2560,0.014556444353527494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,1536,0.009655999640623728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,10240,0.05317688981691996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,2048,0.013812444276279874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,1024,0.007348444726732042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,1536,0.009295110901196798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,1536,0.011809777882364062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,768,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,1024,0.008737777670224508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,1024,0.009324444664849175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,5120,0.019928889142142404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,512,0.0041751112374994485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,768,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,256,0.002532444480392668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,3584,0.014876445134480795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,512,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,128,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,512,0.006606222026877933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,256,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,256,0.006047111004590988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,64,0.0018968888454967076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,128,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,32,0.0021582223061058256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,128,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,2048,0.010993777877754636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,64,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,16384,32,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,65536,0.14560622639126247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,51200,0.18842221630944145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,768,0.00793688909875022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,65536,0.1339093314276801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,16384,0.06464978059132893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,51200,0.11156355010138617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,16384,2560,0.014834667245546976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,51200,0.10780088769065009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,12288,0.05011377731959025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,16384,0.036729776197009616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,10240,0.04249866803487142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,16384,2560,0.012028444144460889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,12288,0.029723554849624634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,8192,0.03558222121662564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,12288,0.04314488834804959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,10240,0.036898665957980685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,7168,0.030806223551432293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,8192,0.021761778328153823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,8192,0.03177244464556376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,65536,0.24179376496209037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,7168,0.019698666201697457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,7168,0.02807377775510152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,5120,0.023763555619451735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,6144,0.0181342214345932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,4096,0.020678222179412842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,5120,0.015993777248594496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,5120,0.021664000219768945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,3584,0.017831110292010836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,4096,0.014131555954615274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,4096,0.019879110985332064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,16384,0.06244977977540758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,3072,0.015818667080667283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,3584,0.013279111021094851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,3584,0.016686222619480558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,2560,0.013785777820481194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,3072,0.012440889245933957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,2048,0.011858666936556498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,2560,0.011226666470368704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,10240,0.025573333104451496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,1536,0.008005333443482717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,2560,0.013745778136783175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,2048,0.009747555686367882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,1024,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,1536,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,2048,0.012792000340090858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,1536,0.010994666980372535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,1024,0.00833511104186376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,6144,0.024614221519894067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,512,0.0031831111345026228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,1024,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,768,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,256,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,768,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,128,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,512,0.007125332951545715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,256,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,512,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,64,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,32,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,12288,128,0.006606222026877933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,256,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,128,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,64,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,65536,0.2043902211719089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,32,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,12288,3072,0.015089778436554803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,51200,0.160060445467631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,16384,0.056997332308027476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,65536,0.13812443945142958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,6144,0.027073777384228174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,12288,0.04235466652446323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,12288,768,0.004982222285535601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,51200,0.09876533349355061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,10240,0.036361777120166354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,16384,0.04567289021280077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,12288,0.030693332354227703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,8192,0.030336889955732558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,12288,0.03751555416319106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,10240,0.02423022190729777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,7168,0.027142221728960674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,10240,0.03174933460023668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,6144,0.02407288882467482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,8192,0.02635555631584591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,7168,0.019159111711714003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,5120,0.02072355482313368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,7168,0.023232888844278123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,6144,0.017204445269372728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,4096,0.017210667332013447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,6144,0.020688888099458482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,5120,0.015531554818153381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,3584,0.015554666519165039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,5120,0.018645332919226754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,4096,0.01540177729394701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,4096,0.016358221570650738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,3072,0.01387555566098955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,3584,0.013117333253224691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,3584,0.014536889062987434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,2560,0.012375999655988483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,51200,0.10711555348502265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,3072,0.012055111428101858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,3072,0.012870222330093384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,2560,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,2560,0.012065778175989786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,1536,0.007722666694058313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,8192,0.02075822154680888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,2048,0.00906755526860555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,1024,0.005664000080691443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,2048,0.01109866632355584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,1536,0.008222222328186035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,768,0.003974222060706881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,1536,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,1024,0.007781333393520779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,1024,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,512,0.0028853333658642242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,768,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,256,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,768,0.006700444552633498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,512,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,128,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,256,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,65536,0.12121689319610596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,64,0.0018977777411540349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,32,0.0018702221827374564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,16384,0.039376888010236956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,128,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,10240,2048,0.009600888523790572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,64,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,65536,0.1702453295389811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,10240,32,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,51200,0.1280719969007704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,16384,0.04341600007481045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,65536,0.06981777482562594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,65536,0.08799999952316284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,51200,0.06896266672346327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,12288,0.034295111894607544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,16384,0.022800000177489385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,10240,0.026704000102149114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,512,0.006803555621041193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,12288,0.028139554791980322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,8192,0.02255200015174018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,10240,0.016295111841625638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,10240,0.023940443992614746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,7168,0.01956266661485036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,10240,128,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,8192,0.014514666464593677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,8192,0.022834666901164587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,6144,0.01715555621518029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,5120,0.014767999450365702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,7168,0.013474666410022311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,7168,0.017794667018784415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,4096,0.012445333103338877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,5120,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,6144,0.01686311099264357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,3584,0.010981332924630908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,5120,0.014476444986131458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,4096,0.011381333072980246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,51200,0.05674488676918877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,3072,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,4096,0.013108444710572561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,3584,0.00964355551534229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,2560,0.008543999658690559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,3584,0.012356444365448423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,2048,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,12288,0.01866311166021559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,2560,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,3072,0.010783111055692038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,1536,0.006200888918505774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,2560,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,1024,0.004509333521127701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,2048,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,2048,0.009681777821646797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,768,0.0031911111954185697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,1024,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,1536,0.007664889097213745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,512,0.002530666689078013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,1024,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,768,0.006320000108745363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,768,0.006543111056089401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,512,0.006220444622966979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,6144,0.012734221915404001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,512,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,128,0.002128888956374592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,256,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,64,0.0019120000716712738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,256,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,128,0.0058551111982928375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,32,0.0018622221218215094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,128,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,65536,0.1469173298941718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,3072,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,64,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,32,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,8192,16384,0.038078222009870745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,51200,0.1159857776429918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,65536,0.08331555790371366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,65536,0.06645688745710585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,8192,1536,0.007140444384680853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,51200,0.0623191131485833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,12288,0.031465778748194374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,16384,0.021726222501860723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,16384,0.03262933426433139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,10240,0.02683199942111969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,8192,256,0.0022693334354294673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,12288,0.017878222796652053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,12288,0.02791644467247857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,8192,0.022612444228596155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,7168,0.017875555488798354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,10240,0.015818667080667283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,8192,0.0138586668504609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,6144,0.01551555593808492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,8192,0.020319999919997323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,7168,0.013096888860066732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,5120,0.013443555268976422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,7168,0.017030222548378837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,6144,0.012137778103351593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,6144,0.016505777835845947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,5120,0.011047999891969891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,5120,0.014507555299335055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,3584,0.010011555420027839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,4096,0.009679999616410997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,4096,0.013110222087966071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,51200,0.053414222266938954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,3072,0.009019555317031013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,3584,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,2560,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,3584,0.011515555282433828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,3072,0.008277333445019191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,2048,0.00676711110605134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,3072,0.010264889233642155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,2560,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,1536,0.005668444352017508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,2560,0.010044444766309526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,2048,0.007290666302045186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,10240,0.021996445126003687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,2048,0.008631111019187504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,1536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,768,0.0032106666929192017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,1536,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,1024,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,512,0.002528000002106031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,1024,0.006565333240562015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,768,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,4096,0.011768889096048145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,256,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,768,0.005994666781690385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,128,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,512,0.005879999862776862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,512,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,256,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,16384,0.03976710968547397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,64,0.0018728888697094387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,7168,128,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,256,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,128,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,65536,0.12078755431705052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,64,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,7168,32,0.005212444398138259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,51200,0.09550666809082031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,16384,0.033667554457982383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,12288,0.026833777626355488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,65536,0.06350666946834989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,65536,0.07565244701173571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,51200,0.05139999919467502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,51200,0.056405332353379994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,16384,0.02095111045572493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,10240,0.022970666488011677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,16384,0.03345333205329047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,7168,1024,0.0038480001191298165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,10240,0.015218666858143277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,12288,0.02459999918937683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,8192,0.013471110827393003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,10240,0.020577778418858845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,6144,0.015464888678656684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,8192,0.018434666925006442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,7168,0.01275111072593265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,5120,0.013739555246300168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,7168,0.016188444362746347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,6144,0.011691555380821228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,4096,0.011037333144081963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,6144,0.01499022212293413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,5120,0.01033866653839747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,3584,0.009471110999584198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,5120,0.013458666702111563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,4096,0.00905511114332411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,3072,0.00832177781396442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,4096,0.01275111072593265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,2560,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,3584,0.010988444089889526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,3072,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,2048,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,3072,0.009845332966910468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,2560,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,2560,0.008742222355471717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,1536,0.00536088893810908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,2048,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,2048,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,8192,0.01956355571746826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,1024,0.0035582222044467926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,7168,0.01735733283890618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,1536,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,768,0.002893333426780171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,1536,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,1024,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,512,0.00249422217408816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,768,0.006228444476922353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,1024,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,256,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,768,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,128,0.001965333397189776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,256,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,512,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,64,0.0018231111268202465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,256,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,6144,32,0.0018639999131361644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,128,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,3584,0.007968000239796108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,128,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,65536,0.10440622435675727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,64,0.00507733349998792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,6144,32,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,51200,0.08272177643246122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,65536,0.060890667968326144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,65536,0.06709777646594577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,51200,0.051028443707360156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,12288,0.01720355616675483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,51200,0.049572444624371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,12288,0.024103111690945093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,16384,0.02016266683737437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,10240,0.02036977807680766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,16384,0.029110223054885864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,12288,0.016767111089494493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,8192,0.01791911158296797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,10240,0.014800000521871777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,7168,0.015841777125994366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,10240,0.019889778561062284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,6144,512,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,8192,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,6144,0.01404800017674764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,8192,0.01866222255759769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,7168,0.012277333272827996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,5120,0.012311110893885294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,6144,0.011362666885058085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,7168,0.01571999986966451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,4096,0.010087110930018956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,6144,0.014486221803559197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,5120,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,3584,0.008663111262851292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,4096,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,3072,0.007819555699825287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,4096,0.012063999970753988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,3584,0.00794222205877304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,2560,0.006899555524190267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,3584,0.011048888994587792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,2048,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,3072,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,3072,0.009368889033794403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,2560,0.007340444458855524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,1536,0.004511111312442356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,2560,0.008009778128729926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,2048,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,1024,0.0032248888164758682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,2048,0.008001777860853408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,768,0.0028568889117903183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,1024,0.006215999937719769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,12288,0.022321777211295232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,512,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,768,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,768,0.005763555566469829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,256,0.0022355556074115965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,512,0.005659555395444234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,512,0.005538666827811136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,128,0.0019022222194406723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,256,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,256,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,5120,0.013122667041089801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,64,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,128,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,128,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,5120,16384,0.030123554997973975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,65536,0.08901066912545098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,64,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,51200,0.06542311112085979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,32,0.005099555684460534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,16384,0.024667556087176006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,65536,0.059577776326073535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,5120,1536,0.00656355544924736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,12288,0.019225777851210702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,1536,0.006973333656787872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,51200,0.046448889705869884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,51200,0.04760711060629951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,5120,1024,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,10240,0.01681866745154063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,16384,0.026421333352724712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,12288,0.016198222835858662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,8192,0.014800000521871777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,12288,0.022135110365019903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,10240,0.01424088908566369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,7168,0.012884444660610624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,10240,0.018758222460746765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,8192,0.012515555653307172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,6144,0.011790222591824003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,8192,0.018457776970333524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,5120,0.01072177787621816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,7168,0.014485332700941296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,6144,0.010713777608341642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,4096,0.00942488925324546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,6144,0.013764444324705334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,5120,0.008842666943868002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,3584,0.010664888554149203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,5120,0.012367111113336353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,4096,0.008150222400824228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,3072,0.009467555416954888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,3584,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,3584,0.009577777650621202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,2560,0.00889955543809467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,3072,0.007652444144090016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,3072,0.00814666681819492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,16384,0.019557333654827543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,2048,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,2560,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,1536,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,2560,0.007240888973077138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,2048,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,1024,0.0042151109211974675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,2048,0.007284444239404466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,1536,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,768,0.0038480001191298165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,7168,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,1536,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,1024,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,512,0.0028524444335036804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,1024,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,768,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,256,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,512,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,768,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,128,0.0018808889306253856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,512,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,64,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,256,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,128,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,4096,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,65536,0.05889066722657946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,128,0.004865777575307422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,65536,0.07645955350663927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,64,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,51200,0.05962577793333265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,32,0.005382222019963794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,65536,0.057849778069390186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,16384,0.022682666778564453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,65536,0.05736266904407077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,12288,0.017549332645204332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,51200,0.04658400019009908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,10240,0.01514400045077006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,16384,0.019105777144432068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,51200,0.04460888769891527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,12288,0.01586844523747762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,16384,0.024949333733982507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,8192,0.013112000293201871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,12288,0.020970667401949566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,10240,0.014141332772043018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,8192,0.01203555530971951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,10240,0.017841777867741056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,6144,0.01089866707722346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,8192,0.01757600075668759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,7168,0.010985777609878115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,5120,0.009918221996890174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,7168,0.014152889450391134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,4096,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,6144,0.009686222506894005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,4096,256,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,6144,0.012855999999576144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,3584,0.009944889280531142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,4096,4096,0.011744000017642975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,5120,0.011233777635627322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,3072,0.008641777767075432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,4096,0.0107004443804423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,3584,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,2560,0.00793422261873881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,3072,0.007635555333561367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,3584,0.008640888664457533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,2048,0.006325333482689328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,3072,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,1536,0.005160000175237656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,2560,0.007247111035717859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,2048,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,2560,0.0069982219073507525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,1024,0.0038622220357259116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,2048,0.0069982219073507525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,1536,0.006555555595291986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,768,0.0035422220826148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,1536,0.006095110956165526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,7168,0.01204622205760744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,1024,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,512,0.0028808888875775864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,768,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,256,0.0021520000365045336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,768,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,512,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,128,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,512,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,256,0.005756444401211209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,64,0.0017893332988023758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,5120,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,256,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3584,32,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,128,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,4096,0.008310221963458592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,65536,0.06143110990524292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,128,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,51200,0.04890666736496819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,64,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3584,32,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,65536,0.05678755707210965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,65536,0.05451644460360209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,51200,0.045871112081739635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,12288,0.01629599928855896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,51200,0.0428071121374766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,10240,0.01349066694577535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,16384,0.01891288823551602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,16384,0.023708444502618577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,12288,0.015761777758598328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,8192,0.011786667009194693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3584,1024,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,10240,0.013745778136783175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,7168,0.010832000109884473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,10240,0.015339555011855232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,8192,0.011499555574523078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,6144,0.0099973330895106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,7168,0.010394666757848527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,8192,0.015156444576051501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,5120,0.009118222528033787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,7168,0.013090666797426013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,6144,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,4096,0.007928888830873702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,6144,0.012031111452314587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,3584,0.007575111256705389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,5120,0.01037688884470198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,4096,0.008269333177142674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,3072,0.008447999755541483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,4096,0.00852177788813909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,2560,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,3584,0.007646222081449296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,16384,0.018935999936527677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,3072,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,2048,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,3072,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,2560,0.007111999723646376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,1536,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,2048,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,2560,0.006917333437336816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,1024,0.0041786668201287585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,2048,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,768,0.0038720000949170855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,1536,0.006585777633719974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,1024,0.006225777582989798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,1536,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,12288,0.018224888377719455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,512,0.0030871110243929755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,256,0.002477333363559511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,768,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,1024,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,128,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,512,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,768,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,64,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,256,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,512,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,3072,32,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,256,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,128,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,3072,5120,0.009022222624884712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,65536,0.05621688895755344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,128,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,51200,0.042928887738121875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,64,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,16384,0.017261332935757108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,32,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,65536,0.05599021911621094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,12288,0.014130666851997375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,65536,0.05208711160553826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,51200,0.045208000474505954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,10240,0.011991110940774282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,51200,0.04138933287726508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,3072,3584,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,8192,0.01038755559258991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,12288,0.01516177753607432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,16384,0.02260622216595544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,7168,0.009714666340086196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,10240,0.01314755529165268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,12288,0.0173653331067827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,6144,0.008737777670224508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,8192,0.01090311093462838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,10240,0.014782221780882942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,8192,0.013749333719412485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,7168,0.010031110710567897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,7168,0.012395555774370829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,6144,0.009571555587980483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,4096,0.007497777541478475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,5120,0.008889777792824639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,6144,0.0105608892109659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,3584,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,5120,0.009341333475377824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,3584,0.00793333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,4096,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,3584,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,2560,0.00701333334048589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,3072,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,3072,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,2048,0.0058764442801475525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,2560,0.00701422244310379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,2560,0.006648889018429651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,1536,0.00564444437623024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,2048,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,2048,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,1024,0.004171555654870139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,1536,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,1536,0.006205333189831839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,5120,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,768,0.0035422220826148987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,1024,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,512,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,768,0.005967999911970562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,1024,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,4096,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,256,0.0024702221983008915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,768,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,128,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,512,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,256,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,512,0.00509066672788726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,3072,0.008211555580298105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,64,0.0017893332988023758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2560,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,128,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,256,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,65536,0.047951999637815684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,128,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,51200,0.036935110886891685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,32,0.005041777673694822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,65536,0.05539110965198941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2560,16384,0.018631999691327412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,65536,0.05048977666431003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,51200,0.044723557101355664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,51200,0.040095110734303795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,16384,0.018173333671357896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,10240,0.010270222193664974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,16384,0.021143999364640977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,12288,0.014138667119873894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,8192,0.009313777916961247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,12288,0.016758221719000075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,10240,0.012144889268610211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,7168,0.008836444053384993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,10240,0.014474666780895658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,6144,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,8192,0.01072088877360026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,7168,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,8192,0.011711111499203576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,5120,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,6144,0.009359111388524374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,7168,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,4096,0.008961777720186446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,6144,0.009356444080670675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,5120,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,3584,0.008676444490750631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,5120,0.008299555215570662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,4096,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,3072,0.005884444548024072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,4096,0.008028444316652086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,3584,0.008113777471913232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2560,64,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,2560,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,16384,0.0141004439857271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,3584,0.007281777759393056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,12288,0.011683555940786997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,2560,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,3072,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,1536,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,1024,0.0038684445122877755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,2560,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,1536,0.006247111078765657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,2048,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,768,0.003567999849716822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,1024,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,512,0.0028648889727062653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,1024,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,768,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,256,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,768,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,512,0.005978666659858491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,512,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,128,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,256,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,128,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,128,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,32,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,2048,2048,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,3072,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,65536,0.03419288992881775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,64,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,2048,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,32,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,65536,0.05462222297986349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,65536,0.04840977655516731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,16384,0.012095111111799875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,2048,1536,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,51200,0.044334222873051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,12288,0.009716444545321995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,51200,0.038733333349227905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,16384,0.016360888878504436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,12288,0.01309511148267322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,16384,0.017714665995703805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,12288,0.014862222803963555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,8192,0.007674666742483775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,10240,0.011921777493423887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,2048,256,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,10240,0.012336889074908363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,7168,0.008068444828192392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,6144,0.007244444555706448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,8192,0.010738666686746808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,8192,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,7168,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,5120,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,7168,0.009400889277458191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,4096,0.006147555592987273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,6144,0.008656889200210571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,5120,0.008831111093362173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,3584,0.005751111027267244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,5120,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,4096,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,51200,0.027152889304690894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,3072,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,4096,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,3584,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,3072,0.007291555404663086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,3584,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,2048,0.005542222410440445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,3072,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,10240,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,1536,0.004581333448489507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,2048,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,2048,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,1024,0.003531555748648114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,1536,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,1536,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,768,0.003165333428316646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,1024,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,1024,0.005373333477311664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,6144,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,768,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,768,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,256,0.002459555657373534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,512,0.00589155571328269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,128,0.0021262222694026097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,512,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,256,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,64,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,2560,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,256,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,128,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,65536,0.023982221881548565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,128,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,51200,0.019655111763212416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1536,2560,0.0069520001610120135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,2560,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,32,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,16384,0.013456000222100152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,65536,0.046784000264273755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,12288,0.010732444624106089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,51200,0.04382222228580051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,51200,0.037556442949506975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,10240,0.009527111219035255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,16384,0.016241777274343703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,12288,0.013129777378506131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,8192,0.009083555804358589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,12288,0.01240799989965227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1536,512,0.002628444383541743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,10240,0.011808888779746162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,7168,0.008485332959228093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,10240,0.011012444065676795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,8192,0.01072355525361167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,6144,0.007611555357774098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,8192,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,7168,0.009771555662155151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,5120,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,7168,0.009295110901196798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,6144,0.009436444275909001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,4096,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,6144,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,5120,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,3584,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,4096,0.008343111309740279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1536,64,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,3072,0.00553955551650789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,65536,0.054120888312657676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,3584,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,2560,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,3584,0.007096000015735626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,3072,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,2048,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,3072,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,2560,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,1536,0.004664888812435998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,2048,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,2048,0.006119110931952794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,1024,0.0038568890757030914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,1536,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,1536,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,768,0.0032177778581778207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,1024,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,1024,0.005445333404673471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,5120,0.008610666626029545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,512,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,768,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,256,0.002193777718477779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,512,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,4096,0.007328000333574083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,128,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,512,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,256,0.005405333307054307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,256,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,64,0.0019822222077184254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,1024,32,0.0019022222194406723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,128,0.005452444569932089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,16384,0.015925332903862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,65536,0.018965333700180054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,128,0.0047155556579430895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,51200,0.015836444165971544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,32,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,65536,0.05397955576578776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,16384,0.013282666603724161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,65536,0.04587022132343716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,51200,0.0437395539548662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,12288,0.00999111102686988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,51200,0.03700088792377048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,16384,0.01566844516330295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,16384,0.014949333336618213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,12288,0.01309866706530253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,8192,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,12288,0.012219555675983429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,10240,0.011738667057620155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,10240,0.011063111325105032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,8192,0.010669333239396414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,1024,768,0.005374222166008419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,6144,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,8192,0.009732444253232744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,7168,0.010047111246320937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,7168,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,5120,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,6144,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,4096,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,6144,0.008728889127572378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,5120,0.008681777450773451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,1024,2560,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,3584,0.005624889085690181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,4096,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,3072,0.0053031109273433685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,5120,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,2560,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,3584,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,3072,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,2048,0.0042071110672420925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,3584,0.007153777612580194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,2560,0.006899555524190267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,1536,0.004577777865860197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,10240,0.008273777862389883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,2560,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,2048,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,1024,0.003560888684458203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,2048,0.006125333408514659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,1536,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,768,0.003208000005947219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,1536,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,7168,0.007257777783605788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,512,0.002521777732504739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,1024,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,768,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,256,0.0021822222818930945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,768,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,512,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,128,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,256,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,512,0.00508355556262864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,256,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,128,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,32,0.001826666709449556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,128,0.004864888886610667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,65536,0.027611555324660406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,64,0.004845333182149463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,4096,0.0075128889746136135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,32,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,51200,0.022692445251676772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,65536,0.051560001240836255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,768,3072,0.0069368887278768755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,16384,0.009122666385438707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,51200,0.038430223862330117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,12288,0.007706666986147563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,16384,0.01555199921131134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,16384,0.014553777045673795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,768,1024,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,10240,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,12288,0.013080000049538083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,8192,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,12288,0.012077333198653327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,10240,0.0120328888297081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,7168,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,10240,0.011015999648306104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,768,64,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,6144,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,8192,0.009667555491129557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,7168,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,7168,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,5120,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,6144,0.009011555876996782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,4096,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,6144,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,5120,0.008683555656009251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,3584,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,5120,0.008278222547637092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,4096,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,65536,0.045151111152437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,3072,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,3584,0.007762666377756331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,51200,0.036553776926464505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,2560,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,3072,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,2048,0.004754666652944353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,3072,0.006590222318967183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,2560,0.0070879997478591064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,1536,0.003546666767862108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,2560,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,2048,0.006591111007663939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,1024,0.0029200000895394217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,2048,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,1536,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,8192,0.01035911093155543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,768,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,1536,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,1024,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,1024,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,768,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,256,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,768,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,512,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,128,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,256,0.0053013331360287136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,64,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,256,0.004587555511130227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,512,128,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,32,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,4096,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,65536,0.01684711045689053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,128,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,3584,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,51200,0.014138667119873894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,64,0.00461422238085005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,65536,0.04644533329539829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,16384,0.006263999889294307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,32,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,12288,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,65536,0.042744888199700244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,51200,0.03735288977622986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,10240,0.005761777775155173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,16384,0.015850666496488783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,51200,0.03456799851523505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,12288,0.013048888908492194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,16384,0.014325333966149224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,10240,0.011794666449228922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,12288,0.01204711116022534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,7168,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,512,512,0.0025128889828920364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,8192,0.01053422192732493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,6144,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,8192,0.00964355551534229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,7168,0.009816889133718278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,5120,0.004588444613748127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,7168,0.009316444396972656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,6144,0.009345778160625035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,4096,0.004534222185611725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,512,512,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,5120,0.008586666650242275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,3584,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,4096,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,5120,0.007701333198282454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,3072,0.004594666676388847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,4096,0.007251555720965068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,3584,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,2560,0.004854222138722737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,3584,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,3072,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,2048,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,3072,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,2560,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,1536,0.0034977777136696708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,2560,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,2048,0.006392000036107169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,1024,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,2048,0.005972444597217772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,8192,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,768,0.0025439999169773525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,1536,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,1024,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,512,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,10240,0.010983111129866706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,1024,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,256,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,768,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,512,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,128,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,256,0.0052968888647026485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,6144,0.008268444074524773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,64,0.0018115554832749898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,256,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,256,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,128,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,65536,0.011694221860832639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,128,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,64,0.004527111020353106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,51200,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,65536,0.04588177800178528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,32,0.004634666774008009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,16384,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,65536,0.041780445310804576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,51200,0.03734488950835334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,12288,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,16384,0.015571556157535978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,256,1536,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,16384,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,10240,0.005304000029961268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,12288,0.013072888884279462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,8192,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,12288,0.012132444315486483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,10240,0.011632888681358762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,7168,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,768,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,8192,0.0107004443804423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,256,512,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,6144,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,8192,0.009649777577983009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,7168,0.00998933364947637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,5120,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,7168,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,6144,0.009356444080670675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,5120,0.008397333323955536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,6144,0.008640000389681922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,3584,0.0046533333758513136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,5120,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,4096,0.00794577764140235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,4096,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,3584,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,2560,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,3584,0.007040888898902469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,3072,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,2048,0.0047884442740016514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,3072,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,2560,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,1536,0.0034959999223550162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,2560,0.006423111177153057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,2048,0.006314666734801398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,1536,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,10240,0.01076533314254549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,1536,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,768,0.002537777854336633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,1024,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,1024,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,4096,0.0042453333735466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,512,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,768,0.005774222314357758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,256,0.0021333334346612296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,512,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,3072,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,128,0.0018702221827374564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,512,0.004621333132187526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,256,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,128,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,32,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,128,0.004947555561860402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,65536,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,64,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,7,128,2048,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,51200,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,16384,0.005530666559934616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,32,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,128,1024,0.002874666617976295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,12288,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,65536,0.04184977875815498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,10240,0.004583111239804162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,51200,0.033548444509506226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,8192,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,16384,0.014468444718254937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,7168,0.004862222406599256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,12288,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,6144,0.004562666846646203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,768,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,10240,0.01074488874938753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,5120,0.00423555572827657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,8192,0.009651555783218807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,4096,0.0039013334446483185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,7168,0.009320889082219865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,3584,0.004232888834344016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,3072,0.0041911109454101985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,6144,0.008390222158696916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,256,0.004678222040335338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,2560,0.004552888787455029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,5120,0.008005333443482717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,2048,0.004226666771703296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,4096,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,1536,0.003536888708670934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,3584,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,1024,0.0028373334142896864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,768,0.002667555585503578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,2560,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,512,0.0023448889454205832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,2048,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,256,0.0021600000974204806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,768,0.005385777602593104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,128,0.0018159999615616268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,512,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,64,0.001818666648533609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,128,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,64,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,64,0.004577777865860197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,65536,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,32,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,51200,0.005977777971161737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,65536,0.04144000013669332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,3072,0.00656355544924736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,51200,0.033915556139416166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,1536,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,12288,0.0052426668504873914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,1024,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,10240,0.004958222309748332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,8192,0.004560889055331548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,16384,0.01408266690042284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,7168,0.004544888933499654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,12288,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,64,256,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,10240,0.010840000377760993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,5120,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,8192,0.009696000152164036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,4096,0.003942222230964237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,7168,0.008974221845467886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,3584,0.00424177779091729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,6144,0.008525333470768398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,3072,0.0038844446341196695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,5120,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,2560,0.004219555606444676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,4096,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,3584,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,1536,0.003519111209445529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,3072,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,128,51200,0.03354933195643955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,1024,0.0028400001012616684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,16384,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,2048,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,512,0.002159111201763153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,1536,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,256,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,1024,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,128,0.001790222194459703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,512,0.00508977762526936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,64,0.001818666648533609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,256,0.004893333133724001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,32,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,6144,0.004518222063779831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,128,0.004514666481150521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,32,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,2048,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,2560,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,16384,0.336900446150038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,7,32,768,0.002507555608948072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,51200,0.4864782227410211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,16384,0.22777512338426378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,12288,0.25599198871188694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,7,32,768,0.005091555416584015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,51200,1.011039098103841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,10240,0.20028977923923066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,16384,0.1371271080440945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,8192,0.16534488730960423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,12288,0.10350933339860703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,8192,0.0715199973848131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,8192,0.11852178308698867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,10240,0.13155999448564318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,6144,0.11414755715264215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,7168,0.13215200106302896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,7168,0.06303466690911187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,6144,0.08902133173412746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,5120,0.09565777911080255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,6144,0.05429333448410034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,4096,0.07986133628421359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,12288,0.15477600362565783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,5120,0.046893331739637584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,5120,0.07323022021187676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,3584,0.06791555881500244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,4096,0.040214220682779946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,10240,0.08598844210306804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,3072,0.05986399783028496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,3584,0.03519999980926514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,3584,0.053005334403779775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,2560,0.05067733261320326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,51200,0.4238293435838487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,3072,0.03155288762516446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,2048,0.04128266705407037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,7168,0.10094400246938069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,3072,0.0503111117415958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,2560,0.027406222290462915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,1536,0.03200888964864943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,2560,0.04320977793799507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,1024,0.022631110416518316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,2048,0.02299733294381036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,1536,0.02068000038464864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,2048,0.03569599986076355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,1536,0.024728000164031982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,1024,0.017927999297777813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,512,0.012649777862760754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,1024,0.0206986665725708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,768,0.01500266624821557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,256,0.00755466686354743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,768,0.014536889062987434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,512,0.012757333616415659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,128,0.004972444640265571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,512,0.013167111410035027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,64,0.004218666503826777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,32,0.0042133331298828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,256,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,4096,0.06811466481950548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,128,0.008030222521887885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,64,0.008811555802822113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,65536,32,0.008669333325492011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,65536,768,0.01773422294192844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,65536,0.5152284304300944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,51200,0.38910667101542157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,51200,0.7672479947408041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,65536,0.9948861863878039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,16384,0.24899199273851183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,128,0.010441777606805166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,12288,0.19058222240871855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,16384,0.18233688672383627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,16384,0.10902221997578938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,10240,0.15956266721089682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,12288,0.12292355961269802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,12288,0.08168800009621514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,65536,0.4252106613583035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,10240,0.10583821932474773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,7168,0.12214222219255234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,10240,0.07009689013163249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,51200,0.3396595584021674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,8192,0.09337066941791111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,8192,0.056926223966810435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,6144,0.1046595573425293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,8192,0.1307848824395074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,5120,0.0887991123729282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,7168,0.05132088727421231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,7168,0.07840889030032687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,6144,0.044974221123589404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,6144,0.07306222120920818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,65536,256,0.010288000106811523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,5120,0.05985599756240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,3584,0.06222666634453667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,5120,0.047260443369547524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,4096,0.03373333480623033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,3072,0.054670221275753446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,4096,0.055925336149003774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,3584,0.029419554604424372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,3584,0.04475377665625679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,2560,0.04584177666240268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,3072,0.025794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,3072,0.03977422250641717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,2560,0.022776888476477727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,1536,0.028502222564485338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,1024,0.020094222492641874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,2048,0.01981511049800449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,2048,0.030737777551015217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,768,0.014928888943460254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,1536,0.01755644381046295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,2560,0.03422044383154975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,512,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,1024,0.01555377741654714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,1024,0.01774488886197408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,768,0.013792888985739814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,256,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,1536,0.020635555187861126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,512,0.010885333021481832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,768,0.013037333057986366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,512,0.011746666497654386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,256,0.009549332989586724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,4096,0.0718640022807651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,64,0.002542222125662698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,256,0.00831822223133511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,32,0.0027964444210131965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,51200,128,0.009330666727489894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,128,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,64,0.008336889247099558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,2048,0.036904888020621404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,65536,0.3182408809661865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,65536,0.17464533117082384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,51200,0.24803378846910265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,16384,0.08460355467266506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,51200,0.13041333357493082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,16384,0.04287199841605293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,51200,0.12424177593655056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,12288,0.06491377618577746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,16384,0.0652551121181912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,51200,128,0.003168888803985384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,10240,0.052990221314960055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,12288,0.033378667301601835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,8192,0.04264444443914625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,12288,0.05317155520121256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,10240,0.028799109988742407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,7168,0.03655111127429538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,8192,0.024361777636739943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,6144,0.03189511100451151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,8192,0.03638133406639099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,7168,0.02221777704026964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,7168,0.03250133328967624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,5120,0.027265777190526325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,6144,0.021747555997636583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,4096,0.023544000254737005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,5120,0.017664889494578045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,65536,0.15449955728318956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,3584,0.019870221614837646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,4096,0.01573866605758667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,4096,0.023012444376945496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,3584,0.014461333552996317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,3584,0.018204443984561496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,2560,0.014828445182906257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,3072,0.013382222089502545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,3072,0.017271111408869427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,2048,0.012149333126015134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,2560,0.012207110722859701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,2560,0.015862221519152325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,10240,0.04348355531692505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,1536,0.00999022192425198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,2048,0.011030221978823343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,1024,0.007336888876226213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,2048,0.014127999544143677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,51200,32,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,768,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,1536,0.011039999624093374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,1024,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,512,0.003940444439649582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,1024,0.009353777600659264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,768,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,6144,0.029003553920321997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,256,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,5120,0.02477333280775282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,128,0.002193777718477779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,512,0.007253333098358578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,512,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,3072,0.01717422240310245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,64,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,128,0.006680888848172293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,16384,32,0.0021840000732077495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,128,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,64,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,65536,0.2410435676574707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,32,0.005744888964626525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,51200,0.18751555018954805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,65536,0.14949244923061794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,1536,0.009730666875839233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,16384,0.06461777951982287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,51200,0.1118497716055976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,65536,0.13454222679138184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,12288,0.04981599913703071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,16384,0.037316444847318865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,16384,0.05610577927695381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,12288,0.029490666257010564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,10240,0.04230933388074239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,51200,0.10639466842015584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,12288,0.0443466669983334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,8192,0.0355004436439938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,10240,0.02570666703912947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,256,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,8192,0.021871111459202234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,10240,0.03698933455679152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,6144,0.027079999446868896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,8192,0.032617777585983276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,7168,0.01994400057527754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,5120,0.023592000206311543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,7168,0.028429332706663344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,6144,0.018976888722843595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,4096,0.020308444897333782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,6144,0.025945777694384258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,5120,0.016210666961140104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,3584,0.017833777599864535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,5120,0.02100533412562476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,4096,0.01424888935354021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,3072,0.015815999772813585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,4096,0.02047644390000237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,3584,0.013348444468445249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,3584,0.017124444246292114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,2560,0.01371022231049008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,3072,0.01239822225438224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,2048,0.011845333708657159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,3072,0.01516711049609714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,2560,0.011360888679822287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,16384,768,0.007017778025733099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,1536,0.008023999631404877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,2560,0.014456889695591398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,2048,0.009669333696365356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,1024,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,16384,256,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,768,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,2048,0.012728888955381183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,1536,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,512,0.0033395555284288195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,1024,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,768,0.007818666597207388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,256,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,512,0.0070053330726093715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,768,0.006855999843941794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,128,0.0021119999388853707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,512,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,256,0.0063884444534778595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,64,0.0018746666610240936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,256,0.006093333164850871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,32,0.0019946667469210094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,64,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,65536,0.20357422033945718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,32,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,51200,0.1606880029042562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,65536,0.13803554905785456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,12288,7168,0.030890666776233252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,16384,0.05674044291178385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,65536,0.12501955032348633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,1024,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,12288,0.042452444632848106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,51200,0.10175377792782253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,51200,0.10708533393012153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,16384,0.035135110219319664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,10240,0.03625244564480252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,8192,0.030255112383100722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,16384,0.04638311266899109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,12288,0.028520888752407495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,12288,0.03829066620932685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,10240,0.024181332853105333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,12288,128,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,7168,0.027057778504159715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,6144,0.023956444528367784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,8192,0.020784000555674236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,10240,0.03188800149493747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,5120,0.02055111030737559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,7168,0.018927110566033255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,4096,0.017595556047227647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,6144,0.018304889400800068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,7168,0.023976888921525743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,6144,0.021369778447681006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,3584,0.01570311188697815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,5120,0.015454222758611044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,3072,0.013815999858909182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,5120,0.01847111185391744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,4096,0.01370133293999566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,2560,0.01237155579858356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,3584,0.013064888616402945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,4096,0.016561778055297006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,3072,0.011958222422334882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,3584,0.014568888478808932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,2048,0.01036622209681405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,2560,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,1536,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,12288,1536,0.010292444792058732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,2560,0.012627555264366997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,1024,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,1536,0.00831822223133511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,2048,0.011357333097192975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,768,0.003903999924659729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,1536,0.008734222087595198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,512,0.003035555490189128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,1024,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,768,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,256,0.0023617777559492323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,768,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,512,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,128,0.0021235555824306277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,512,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,256,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,8192,0.02804533309406704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,64,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,256,0.005613333235184352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,10240,32,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,128,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,128,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,65536,0.16775643825531006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,64,0.005680888891220093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,32,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,65536,0.09228710995780097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,51200,0.12740888860490587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,10240,3072,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,65536,0.07250044743220012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,16384,0.043266667260064014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,2048,0.009285333255926767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,12288,0.034225778447257146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,51200,0.06867288880878024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,51200,0.055604444609748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,16384,0.02297244469324748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,10240,0.026692445079485577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,10240,1024,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,16384,0.0348364445898268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,8192,0.023119999302758112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,12288,0.018328888548745047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,12288,0.02865333358446757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,7168,0.01959288948112064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,10240,0.016598222984208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,6144,0.01724444495307075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,8192,0.014550222290886773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,10240,0.026955554882685345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,5120,0.014833778142929077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,7168,0.013402666482660504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,8192,0.020946666598320007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,4096,0.012508444488048553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,6144,0.013117333253224691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,3584,0.011074666347768573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,5120,0.011696889168686338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,4096,0.010715555813577441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,5120,0.015463110473420886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,3072,0.01000533335738712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,4096,0.013120888835854001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,2560,0.008642666869693333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,3584,0.011745777395036487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,3072,0.008984000318580205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,2048,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,3072,0.011031111081441244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,2560,0.008047111332416534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,1536,0.006162666612201267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,2560,0.010426667001512317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,2048,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,1024,0.004582222137186262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,2048,0.009477333062224919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,1536,0.007151111132568783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,768,0.0031991110493739447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,1536,0.0069777775141927935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,1024,0.0069271110826068455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,512,0.00272088891102208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,1024,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,768,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,256,0.0021804444905784395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,768,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,512,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,7168,0.01893422173129188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,128,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,512,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,6144,0.01651022169325087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,256,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,64,0.0018622221218215094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,8192,32,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,256,0.0052737775776121355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,3584,0.010664000279373594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,128,0.00498488876554701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,65536,0.14664622147878012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,64,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,8192,32,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,51200,0.1160000032848782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,65536,0.08095110787285699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,65536,0.06558577881919013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,12288,0.03146222233772278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,51200,0.062423110008239746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,16384,0.021928000781271193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,10240,0.02665422194533878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,12288,0.017786666750907898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,8192,0.022464000516467627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,12288,0.026707554856936138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,10240,0.015817777978049386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,10240,0.023331556055280898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,7168,0.017858665850427415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,8192,0.013956444131003486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,6144,0.015506666567590503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,8192,0.020670221911536325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,7168,0.013069333301650153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,5120,0.013431111143694984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,7168,0.01753155555990007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,6144,0.012101333174440594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,4096,0.011718221836619906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,6144,0.01644000079896715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,5120,0.011060444845093621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,3584,0.010091555615266165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,16384,0.039444443252351545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,5120,0.014165333575672574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,4096,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,3072,0.009191110730171204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,4096,0.012758221891191272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,3584,0.008980444735950893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,51200,0.05332977904213799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,3072,0.00831288927131229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,3584,0.011391999820868174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,3072,0.010880889164076911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,16384,0.03355022271474203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,2560,0.00755466686354743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,1536,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,2560,0.010088000032636855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,2048,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,1024,0.003882666842805015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,2048,0.008606221940782335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,1536,0.0069146665434042616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,768,0.0028906667398081887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,1536,0.006689777804745569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,1024,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,512,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,768,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,1024,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,256,0.0022213332768943575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,768,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,512,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,128,0.001959111127588484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,512,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,256,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,64,0.001812444378932317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,256,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,7168,128,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,128,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,2560,0.007961778177155389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,64,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,2048,0.006801777829726537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,65536,0.1208382182651096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,7168,32,0.0051893335249688895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,51200,0.09558755821651882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,16384,0.03365688853793674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,65536,0.0751511123445299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,65536,0.06338844696680705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,51200,0.0512951115767161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,12288,0.026663111315833196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,51200,0.05624533361858792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,16384,0.02088711162408193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,12288,0.01724799970785777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,8192,128,0.005899555567238066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,8192,0.019726221760114033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,12288,0.024150222539901733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,7168,0.01745244363943736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,10240,0.015307555596033732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,10240,0.02180533276663886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,6144,0.015525332755512662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,8192,0.013423110875818463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,8192,0.019467555814319186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,7168,0.012636444634861417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,5120,0.01364177796575758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,7168,32,0.0018800000349680583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,4096,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,6144,0.011690666278203329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,5120,0.010464888479974536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,6144,0.015836444165971544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,3584,0.009388444324334463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,4096,0.009331555830107795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,5120,0.013424889081054263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,3072,0.008375999828179678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,4096,0.012880889077981314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,3584,0.008657777474986183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,2560,0.0075093333919843035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,3584,0.010957332948843638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,3072,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,2048,0.0063075555695427795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,3072,0.01037333326207267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,2560,0.00741244438621733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,1536,0.0053022222386466126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,2560,0.009660444325870937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,10240,0.022996443841192458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,2048,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,1536,0.006606222026877933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,2048,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,16384,0.030382222599453394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,768,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,512,0.002510222295920054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,1024,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,768,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,256,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,768,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,512,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,128,0.0018639999131361644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,512,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,256,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,64,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,32,0.0019004444281260173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,256,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,6144,128,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,128,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,65536,0.10433599683973525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,64,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,32,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,51200,0.08290755748748779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,65536,0.06636266575919257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,65536,0.06073511309093899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,16384,0.02999911043379042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,51200,0.05056977934307522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,12288,0.024168888727823894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,6144,1024,0.0034497777620951333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,16384,0.020264888803164165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,10240,0.020280000236299302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,1536,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,12288,0.016672000288963318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,8192,0.01789955629242791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,10240,0.01481066644191742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,10240,0.021299555897712708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,7168,0.015789333317014907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,6144,7168,0.016519111063745286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,8192,0.018512000640233357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,6144,0.014144000079896716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,7168,0.012291555603345236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,7168,0.01584533353646596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,5120,0.012331555287043253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,6144,0.011329777538776398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,6144,0.01568888955646091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,4096,0.009647111097971598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,5120,0.009892444643709395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,51200,0.04929688904020521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,5120,0.013086222112178802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,3584,0.008636444807052612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,4096,0.008328888979223039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,4096,0.012728888955381183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,3072,0.007703111403518253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,3584,0.008300444318188561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,2560,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,12288,0.023934221929974023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,3584,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,3072,0.007622222105662028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,2048,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,3072,0.00964355551534229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,2560,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,1536,0.004543999830881755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,2560,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,2048,0.0069724445541699724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,1024,0.003162666741344664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,8192,0.01331999980741077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,2048,0.007720000214046902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,768,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,1536,0.00639111093348927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,512,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,768,0.006227555374304454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,768,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,256,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,512,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,512,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,128,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,256,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,256,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,64,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,128,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,16384,0.026943110757403906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,5120,32,0.0018791111393107309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,128,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,65536,0.08873422278298272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,64,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,51200,0.06508088774151273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,32,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,65536,0.06000889009899563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,16384,0.02462755474779341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,1536,0.006607111129495833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,12288,0.019301333361201815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,51200,0.04754844307899475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,5120,1024,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,16384,0.01941422290272183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,10240,0.01683466633160909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,5120,1024,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,8192,0.014579556054539151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,10240,0.014431110686726041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,12288,0.02239377796649933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,7168,0.012995555169052549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,10240,0.019115555617544387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,8192,0.012719111310111152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,8192,0.017519111434618633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,7168,0.011676444775528379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,7168,0.015119110544522604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,5120,0.010660444696744284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,6144,0.010328888893127441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,6144,0.014630221658282809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,4096,0.009424000150627559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,5120,0.00903911143541336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,5120,0.012355555262830524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,65536,0.058818664815690785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,3584,0.010656889114114972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,4096,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,4096,0.011394666300879585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,3072,0.008943110704421997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,51200,0.046421332491768726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,3584,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,2560,0.008536000218656328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,16384,0.026349332597520616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,3072,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,2048,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,12288,0.016198222835858662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,1536,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,3072,0.008379555410808986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,2560,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,2048,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,1024,0.0042408886882993905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,1536,0.0064604443808396654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,2048,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,6144,0.011882666912343768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,768,0.003531555748648114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,1536,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,512,0.0031902222997612427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,1024,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,768,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,256,0.0021395554973019496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,768,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,512,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,128,0.0018079999006456798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,256,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,256,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,128,0.005590222362014983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,128,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,4096,32,0.0018604445374674266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,65536,0.07629066705703735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,3584,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,64,0.00510044437315729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,2560,0.007103111180994246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,32,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,51200,0.059522668520609535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,16384,0.022639999787012737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,65536,0.05721600188149346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,51200,0.04661422305636936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,12288,0.01756177842617035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,51200,0.04452711012628344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,16384,0.0192266669538286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,10240,0.015093333191341825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,16384,0.025204444924990337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,12288,0.015789333317014907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,8192,0.013041777743233575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,12288,0.020241777102152508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,4096,512,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,10240,0.01425066590309143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,7168,0.012091555529170565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,10240,0.0184568895234002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,8192,0.012171555724408893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,6144,0.01092355532778634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,7168,0.01110577748881446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,8192,0.016413332687483895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,5120,0.009844444692134857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,6144,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,7168,0.014000889327791003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,4096,0.008641777767075432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,6144,0.012724444270133972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,3584,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,4096,0.008307555483447181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,5120,0.011392888923486074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,3072,0.008607111043400234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,4096,0.010056888891590966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,65536,0.05788889196183947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,3584,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,2560,0.007631111476156447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,3584,0.008275555239783393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,2048,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,4096,1024,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,3072,0.007722666694058313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,2560,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,1536,0.005218666460778978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,2560,0.007705777883529663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,1024,0.004176889028814104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,2048,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,2048,0.006638222270541721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,1536,0.0064080001579390625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,768,0.0034462221794658234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,512,0.002826666666401757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,1536,0.00598933340774642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,1024,0.005991111199061076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,768,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,256,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,1024,0.006210666563775804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,128,0.0019680000841617584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,512,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,256,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,512,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,64,0.001875555556681421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,256,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,128,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,5120,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,128,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,65536,0.06080444653828939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,51200,0.04863377743297153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,64,0.004933333231343163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,32,0.004859555512666702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,16384,0.019056000643306308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,65536,0.05682222048441569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,65536,0.054690665668911405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,51200,0.04576711191071404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,12288,0.015604444675975375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,51200,0.042677332957585655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,16384,0.01909244391653273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,10240,0.013463999662134381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,12288,0.015844444433848064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,16384,0.02363377809524536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,8192,0.011772444678677453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,12288,0.019846222466892667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,7168,0.010746666954623329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,8192,0.011402666568756104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,10240,0.016918222109476726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,8192,0.015056000815497505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,6144,0.01000711073478063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3584,768,0.005396444350481033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,7168,0.012440889245933957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3584,32,0.0018666666001081467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,6144,0.009326222042242685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,6144,0.012359999948077731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,4096,0.00794488853878445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,5120,0.008751110898123847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,5120,0.01015644437736935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,3584,0.007583111524581909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,4096,0.00811822215716044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,4096,0.009032888544930352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,3072,0.008220444122950235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,3584,0.007663999994595845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,3584,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,2560,0.007356444166766272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,10240,0.013781333135233985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,3072,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,2048,0.005993777679072485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,2560,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,2560,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,7168,0.010332444475756751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,2048,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,2048,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,1024,0.004265777766704559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,5120,0.009208888643317753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,1536,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,768,0.0038862224254343244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,1536,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,1024,0.006303999986913469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,512,0.0029564443975687027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,1024,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,768,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,256,0.0024684444069862366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,768,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,512,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3584,3072,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,128,0.002238222294383579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,512,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,256,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,256,0.005043555465009478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,64,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,32,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,3072,0.007335111498832703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,128,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,65536,0.05563733312818739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,51200,0.04271466533342997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,64,0.005065777649482091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,3072,32,0.004949333353175057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,16384,0.01719555589887831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,65536,0.051995555559794106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,3072,1536,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,12288,0.01367111090156767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,51200,0.04523644513554043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,51200,0.041307555304633245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,16384,0.01887022289964888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,10240,0.012080888781282636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,16384,0.02130844526820713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,12288,0.01513066722287072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,8192,0.010471999645233154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,12288,0.017857778403494094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,10240,0.013091555900043912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,7168,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,10240,0.016136889656384785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,6144,0.008762666748629676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,8192,0.011023110813564725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,8192,0.013669333524174161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,5120,0.007983999947706858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,7168,0.010039110978444418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,4096,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,6144,0.009512888888518015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,7168,0.012366222010718452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,6144,0.011368888947698804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,5120,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,3584,0.006911110960774952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,3072,0.0074826669361856245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,3072,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,5120,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,2560,0.007012444237867991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,3584,0.00797155582242542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,4096,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,3072,0.007263999846246507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,2048,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,3072,0.007258666886223688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,2560,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,1536,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,65536,0.0561351113849216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,2048,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,1024,0.004202666795916027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,1536,0.006616888774765863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,768,0.003547555456558863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,1024,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,1536,0.005981333139869902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,512,0.00282222218811512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,1024,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,768,0.005980444451173146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,256,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,768,0.005356444252861871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,512,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,512,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,128,0.0021537778278191886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,256,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,256,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,64,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,128,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2560,32,0.0018088887963030075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2560,4096,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,128,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,65536,0.04877955714861552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,3584,0.007655111451943715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,64,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,51200,0.03664266732003953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,32,0.004861333303981357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,16384,0.014166222678290473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,2560,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,65536,0.05519288778305054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2560,2048,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,51200,0.044518222411473594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,51200,0.040072888135910034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,16384,0.018244443668259513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,10240,0.010290666586822933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,12288,0.013752889302041797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,16384,0.020426667398876615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,8192,0.0094355551732911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,12288,0.016975111431545682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,10240,0.012774222426944308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,7168,0.008897778060701158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,10240,0.01514933341079288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,8192,0.010727110836240979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,6144,0.007985778152942657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,8192,0.012814222110642327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,5120,0.007289778027269576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,7168,0.00997688869635264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,6144,0.009366222553782994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,4096,0.009154666629102495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,6144,0.009323555562231276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,5120,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,3584,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,5120,0.008343111309740279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,4096,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,3072,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,4096,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,3584,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,65536,0.050139556328455605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,2560,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,3072,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,12288,0.01144177797767851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,2048,0.006534222099516127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,3072,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,1536,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,2560,0.0070053330726093715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,2048,0.006973333656787872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,2560,0.006610666712125142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,1536,0.006236444330877728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,2048,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,768,0.0035297779573334586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,1536,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,1024,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,1024,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,7168,0.009993777506881291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,768,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,256,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,768,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,512,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,128,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,256,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,64,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,256,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,32,0.0019111111760139465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,128,0.004688888788223267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,65536,0.03398133317629496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,3584,0.007344000041484833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,64,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,51200,0.027170666389995154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,32,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,65536,0.05473600162400139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,16384,0.012088889049159156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,65536,0.04821244544453091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,1024,0.003879111260175705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,51200,0.044191999567879565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,12288,0.009711999860074785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,51200,0.038457777765062116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,16384,0.01721155477894677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,10240,0.008753777378135258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,12288,0.013162666724787818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,16384,0.0177208897140291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,8192,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,2048,512,0.0028817777832349143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,12288,0.014765333798196582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,7168,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,8192,0.010713777608341642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,10240,0.01254488858911726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,6144,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,8192,0.009986666341622671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,2048,512,0.0052879999081293745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,5120,0.006225777582989798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,7168,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,6144,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,4096,0.006213333457708359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,6144,0.008767111433876885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,5120,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,3584,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,4096,0.007633777956167857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,3584,0.007673777639865875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,3584,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,3072,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,3072,0.007333333293596904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,2560,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,3072,0.007158222297827403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,10240,0.012067555553383298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,2048,0.005474666754404704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,2560,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,7168,0.01000088867213991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,2560,0.006582222051090664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,1536,0.0047608887155850725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,2048,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,2048,128,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,1024,0.0038897775941424896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,2048,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,1536,0.006249777972698212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,768,0.0032071111102898917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,1536,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,512,0.002512000087234709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,4096,0.00798044436507755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,1024,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,256,0.0024666666156715816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,768,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,512,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,128,0.002124444478087955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,512,0.005224888937340842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,256,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,64,0.0018764444523387486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,256,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,128,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1536,32,0.0018622221218215094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,128,0.0047884442740016514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,65536,0.023815999428431194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,64,0.004857777721352047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,51200,0.019546666079097323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1536,32,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,65536,0.046799110041724316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,16384,0.013117333253224691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,51200,0.043846223089430064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,51200,0.03737511237462362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,5120,0.008996444443861643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,12288,0.011188444164064197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,1024,0.006238222122192383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,16384,0.015843555331230164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,10240,0.009730666875839233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1536,768,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,8192,0.009154666629102495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,12288,0.013080000049538083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,12288,0.012235555383894177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,8192,0.01034222212102678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,10240,0.01105955574247572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,6144,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,8192,0.010039110978444418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,7168,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,5120,0.006596444381607904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,7168,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,6144,0.009358222285906473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,4096,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,6144,0.008983111215962304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,5120,0.009133333133326637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,3584,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,5120,0.008011555506123437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,4096,0.008025777836640676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,65536,0.05455199877421061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,3072,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,4096,0.007419555551475949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,2560,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,3584,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,3072,0.007599111232492659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,3584,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,2048,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,3072,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,2560,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,1536,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,16384,0.016027554869651794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,2560,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,2048,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,1024,0.0038897775941424896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,768,0.003859555555714501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,2048,0.006292444550328785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,1024,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,7168,0.008623110751310984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,1536,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,512,0.002603555512097147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,256,0.0021786666992637846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,768,0.005945777727497949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,1024,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,128,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,768,0.00527555536892679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,512,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,64,0.0018577778504954444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,256,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,512,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,1024,32,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,128,0.0052693333062860705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,256,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,65536,0.018912000788582694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,128,0.004621333132187526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,64,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,51200,0.015800888339678448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,65536,0.05378577775425381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,1024,32,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,16384,0.013453333742088742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,65536,0.04595288965437147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,51200,0.04313333498107063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,12288,0.009945777555306753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,51200,0.03718577822049459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,16384,0.015850666496488783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,10240,0.008223111430803934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,12288,0.013408888545301227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,16384,0.0151582227812873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,8192,0.007608888877762689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,12288,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,10240,0.011720889144473605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,7168,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,1536,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,8192,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,10240,0.011061333119869232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,6144,0.00664444433318244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,1024,10240,0.011704000333944956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,8192,0.009685333404276106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,5120,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,7168,0.010040000081062317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,4096,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,6144,0.009006222089131674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,7168,0.009312888814343346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,3584,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,5120,0.009003555609120263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,6144,0.008948444492287105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,4096,0.008294222255547842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,5120,0.00796088907453749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,4096,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,2560,0.00499288903342353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,3072,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,3584,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,2048,0.0042151109211974675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,3072,0.0069102222720781965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,1536,0.004580444345871608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,2560,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,2048,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,1024,0.0035617777870761026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,2048,0.006072000082996156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,1536,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,768,0.003077333379122946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,1536,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,1024,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,512,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,1024,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,768,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,256,0.002321777865290642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,512,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,768,0.005305777821275923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,128,0.001989333372977045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,512,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,256,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,64,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,32,0.0018195555441909367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,256,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,65536,0.026707554856936138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,128,0.00462311092350218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,64,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,3584,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,768,32,0.004531555705600315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,51200,0.021892444954978094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,65536,0.04757600029309591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,16384,0.009361777868535783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,65536,0.045406222343444824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,2560,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,12288,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,51200,0.03836266530884637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,10240,0.0069226668112807805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,16384,0.015821332732836407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,51200,0.03658755620320638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,12288,0.013056889176368713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,8192,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,16384,0.014463111758232117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,7168,0.005285333428117964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,12288,0.012079111403889127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,10240,0.01204355557759603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,6144,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,8192,0.010384000009960597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,10240,0.011037333144081963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,5120,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,7168,0.009694221946928237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,8192,0.009904888768990835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,4096,0.005963555640644497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,7168,0.009130666653315226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,5120,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,6144,0.008658666577604082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,3584,0.005570666657553778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,5120,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,4096,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,3072,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,4096,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,3584,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,768,3072,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,3584,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,3072,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,2048,0.004858666823969947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,3072,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,2560,0.007083555890454187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,1536,0.0035031110876136353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,2560,0.006313777632183499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,2048,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,1024,0.0029111111329661477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,2048,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,1536,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,768,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,1536,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,1024,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,512,0.002472888885272874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,1024,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,768,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,256,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,6144,0.009111999637550777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,768,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,128,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,512,0.005029333134492238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,256,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,64,0.001822222231162919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,768,128,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,128,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,32,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,128,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,65536,0.016536888149049547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,512,2560,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,51200,0.014161777165200977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,64,0.004584889031118817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,32,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,16384,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,65536,0.04279111160172356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,51200,0.0370435549153222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,51200,0.034056001239352755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,12288,0.004956444518433677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,16384,0.015836444165971544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,10240,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,16384,0.014460444450378418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,12288,0.012807110945383707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,8192,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,12288,0.011907555162906647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,10240,0.011714666253990598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,7168,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,8192,0.01070666644308302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,10240,0.011045333411958484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,6144,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,512,512,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,8192,0.009704888694816166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,5120,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,7168,0.009123555488056606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,6144,0.009350222018029954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,4096,0.004530666602982415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,5120,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,6144,0.008712000317043727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,512,256,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,3584,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,5120,0.007663999994595845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,4096,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,3584,0.007625777688291337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,4096,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,2560,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,3072,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,3584,0.007296889192528195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,2048,0.00452888881166776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,3072,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,2560,0.0070017774899800616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,2048,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,2560,0.006423111177153057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,1024,0.002862222285734283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,2048,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,1536,0.006136888845099344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,768,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,1024,0.006167111297448476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,1024,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,512,0.002472888885272874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,768,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,768,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,7168,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,256,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,512,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,128,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,256,0.005386666705211003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,3072,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,256,0.004839111119508743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,64,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,32,0.0018177777528762817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,128,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,65536,0.04630577895376417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,65536,0.011717333561844297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,64,0.004763555609517627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,256,1536,0.003536000019974179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,51200,0.009876444107956356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,65536,0.045885334412256874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,16384,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,51200,0.037279999918407865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,1536,0.006008000009589725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,51200,0.033422221740086876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,12288,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,16384,0.015798222687509324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,16384,0.014129777749379476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,10240,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,12288,0.013065777719020844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,12288,0.011807999677128263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,8192,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,512,0.0052382221652401825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,10240,0.011919111013412476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,10240,0.0107360002067354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,7168,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,8192,0.010670222342014313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,256,128,0.0052800000541739995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,6144,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,8192,0.00941244430012173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,7168,0.009992889232105678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,5120,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,6144,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,7168,0.008967999782827165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,4096,0.0042453333735466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,256,32,0.004567111117972268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,5120,0.008655110994974772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,3584,0.004620444443490771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,5120,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,4096,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,3072,0.00452888881166776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,65536,0.04176266656981575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,3584,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,4096,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,2560,0.004527111020353106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,2048,0.004187555362780889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,3584,0.007033777733643849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,3072,0.007319110963079665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,2560,0.0069573331210348345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,1536,0.003504000190231535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,3072,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,1024,0.0028595555987623003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,2048,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,2560,0.006262222097979651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,768,0.0025395556456512874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,2048,0.006031111296680238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,1536,0.006561777657932705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,512,0.002489777695801523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,1024,0.005927111125654644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,1536,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,768,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,256,0.0021226666867733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,512,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,768,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,128,0.001818666648533609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,256,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,512,0.004987555659479565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,64,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,128,32,0.0018168888572189545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,6,128,128,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,256,0.0047022220161226065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,65536,0.006960888703664144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,51200,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,128,0.004586666822433472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,16384,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,6144,0.008164444731341468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,64,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,12288,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,32,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,10240,0.004588444613748127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,8192,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,65536,0.04147911071777344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,7168,0.004550222307443619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,51200,0.03342666559749179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,6144,0.004576888763242298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,12288,0.012054222325483957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,5120,0.004541333350870344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,8192,0.009488888912730748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,4096,0.0041831110914548235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,7168,0.008964444200197855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,3584,0.004522666749027041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,3072,0.0041964443193541635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,6144,0.008359111017651027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,2560,0.0042275554604000514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,5120,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,2048,0.0041857775714662345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,128,1024,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,4096,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,3584,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,1024,0.002817777709828483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,3072,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,768,0.0024951110697454875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,2560,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,512,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,2048,0.006248888870080312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,256,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,1024,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,128,0.001810666587617662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,768,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,64,0.0016168888865245713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,512,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,32,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,256,0.004978666702906291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,128,0.004567111117972268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,16384,0.014119111829333834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,65536,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,64,0.004610666798220741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,51200,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,16384,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,32,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,12288,0.005223111146026188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,65536,0.0414266652531094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,51200,0.03338399860594008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,10240,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,16384,0.014105778601434497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,8192,0.00425511101881663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,7168,0.004541333350870344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,64,1536,0.003499555504984326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,12288,0.012058667010731168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,6144,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,5120,0.004574222283230888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,10240,0.010684444672531553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,4096,0.0038515557017591265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,8192,0.009353777600659264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,1536,0.005923555543025334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,7168,0.009024889104896123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,6144,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,3072,0.003899555653333664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,2560,0.0041831110914548235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,5120,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,2048,0.004229333251714706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,4096,0.007027555671003129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,1536,0.003495111233658261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,3584,0.007334222396214803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,3072,0.0069368887278768755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,2560,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,768,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,2048,0.006270222365856171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,512,0.0022897778285874263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,1536,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,256,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,1024,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,128,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,768,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,64,0.0014942222171359593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,512,0.004975111120276981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,32,0.0017911110901170308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,256,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,128,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,64,0.00453955555955569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,32,32,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,3584,0.00424799985355801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,16384,0.33760889371236164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,6,32,1024,0.0028773333049482773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,16384,0.18506222301059297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,51200,0.48543469111124676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,6,64,10240,0.010670222342014313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,12288,0.257173326280382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,10240,0.21326933966742623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,12288,0.15373155805799696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,12288,0.1043635540538364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,16384,0.13709155718485513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,8192,0.16464266512129042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,10240,0.08697244193818833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,10240,0.12939288881089953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,7168,0.14410489135318333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,6144,0.11352354950375027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,8192,0.07165333297517565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,8192,0.12132444646623398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,51200,0.43673332532246906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,7168,0.07454133033752441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,5120,0.09530399905310737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,7168,0.10260711113611858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,4096,0.08027911186218262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,6144,0.0872560011015998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,3584,0.06921066840489705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,5120,0.07676444451014201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,5120,0.04645244280497233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,4096,0.06283022297753228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,3072,0.060397333568996854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,3584,0.035291555855009295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,2560,0.05138577686415779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,51200,1.0792311562432182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,2048,0.04095288780000474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,3072,0.031496001614464655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,3072,0.049851556619008384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,1536,0.03172889020707872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,2560,0.026891556051042344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,2560,0.03871377640300327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,2048,0.023319111929999456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,1024,0.022505778405401442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,1536,0.022421333524915908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,1024,0.01790577835506863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,768,0.017481777403089736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,1536,0.02791911032464769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,1024,0.0206995556751887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,512,0.01256533298227522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,2048,0.03580088747872247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,256,0.007621333003044128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,6144,0.054758220911026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,768,0.016499555773205228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,512,0.012786666552225748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,128,0.00499555551343494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,768,0.015319999721315173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,4096,0.04902133345603943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,64,0.0041857775714662345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,256,0.010216000179449717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,128,0.01035733355416192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,65536,32,0.0041866666740841335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,3584,0.05164622267087301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,128,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,64,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,32,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,65536,0.514445357852512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,51200,0.3886391056908502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,51200,0.7882497575547961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,65536,1.0086409250895183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,65536,256,0.010425777898894416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,16384,0.2569199932946099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,12288,0.1892684433195326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,16384,0.11083733373218113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,65536,0.46703020731608075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,12288,0.12449867195553249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,16384,0.1716195609834459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,12288,0.08292444547017415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,10240,0.1580853329764472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,8192,0.1295671065648397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,10240,0.10347466336356269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,51200,0.341306659910414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,10240,0.07047733333375719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,7168,0.11348977353837754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,65536,512,0.013223110801643796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,6144,0.10432089037365383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,8192,0.09331644243664211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,7168,0.05012977785534329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,5120,0.08833955393897162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,4096,0.07152533531188965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,6144,0.07134933604134454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,5120,0.0647546648979187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,5120,0.03780622283617655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,6144,0.05551200111707052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,3072,0.05496533380614387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,3584,0.028743998871909246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,3584,0.04151733385192023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,2560,0.045349333021375865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,3072,0.02569244470861223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,3072,0.04197955462667677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,2048,0.036622222926881574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,2560,0.022493332624435425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,8192,0.05799733267890083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,2048,0.029927998781204224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,2048,0.019374221563339233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,7168,0.08209777540630765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,1536,0.02860266632503933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,1024,0.020039111375808716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,1536,0.01891288823551602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,1536,0.023367111881573994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,3584,0.06273333231608073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,1024,0.016878222425778706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,512,0.010346666806273991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,4096,0.0566577778922187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,768,0.013487111363146039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,256,0.006321777900060018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,768,0.014448000325096978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,512,0.011206222077210745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,128,0.003206222214632564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,256,0.009364444348547194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,256,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,64,0.002532444480392668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,128,0.009482666850090027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,2560,0.033293333318498396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,32,0.002830222249031067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,128,0.007268444531493717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,32,0.007667555577225155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,65536,0.3182071050008138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,51200,768,0.014854222536087036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,65536,0.17206222481197783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,51200,0.24732799000210234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,65536,0.1550017727745904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,51200,0.1300977733400133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,512,0.011585777832402123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,16384,0.084897776444753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,12288,0.06430577569537692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,51200,0.12456177340613471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,16384,0.04172711239920723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,10240,0.05282488796446058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,16384,0.06577510966195001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,12288,0.049518220954471164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,8192,0.0414959987004598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,51200,64,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,7168,0.03631022241380479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,10240,0.043375998735427856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,8192,0.024656888511445787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,8192,0.03691111008326212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,6144,0.031253334548738264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,1024,0.015483554866578845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,5120,0.027388445205158655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,7168,0.032736887534459434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,51200,4096,0.03263555632697211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,7168,0.021935110290845234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,6144,0.02895555562443203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,4096,0.022993778189023335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,5120,0.02181422213713328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,3584,0.019890666007995605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,5120,0.025028443998760645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,4096,0.015615110596021017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,3072,0.01706755492422316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,4096,0.02216533323129018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,3584,0.014316444595654806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,2560,0.014762666490342883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,3584,0.01806400054030948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,3072,0.01331911070479287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,12288,0.03378399875428941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,2048,0.012280888855457306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,3072,0.018182223041852314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,2560,0.011900444825490316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,1536,0.009647111097971598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,2560,0.0148062225845125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,1024,0.007576000359323289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,1536,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,2048,0.013756444056828817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,768,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,1536,0.011624889241324531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,1024,0.008966222405433655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,512,0.0041368889311949415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,1024,0.009336888790130615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,768,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,256,0.002521777732504739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,768,0.007683555285135905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,6144,0.019925332731670804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,512,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,128,0.0021911110315057966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,256,0.006574222197135289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,64,0.002096000024014049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,128,0.006653333289755716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,256,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,16384,32,0.0021582223061058256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,128,0.005504000104135937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,64,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,32,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,65536,0.2413946787516276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,2048,0.01072266697883606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,65536,0.14554311169518366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,16384,0.06452622016270955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,65536,0.1350462171766493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,51200,0.11143110858069526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,51200,0.10968533489439224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,12288,0.04994044370121426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,16384,0.05302933189604017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,10240,0.04230044285456339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,12288,0.02993511160214742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,12288,0.04646133383115133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,10240,0.025787555509143408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,16384,512,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,10240,0.0362337761455112
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,8192,0.02163733376397027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,8192,0.03206488821241591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,6144,0.026964444253179762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,7168,0.028505778974956934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,6144,0.022557333111763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,51200,0.18696267075008818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,6144,0.02527644402450985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,5120,0.02358666724628872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,4096,0.020655110478401184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,5120,0.01885422236389584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,16384,0.037235554721620344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,5120,0.0225582222143809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,3584,0.01789422167672051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,4096,0.016892444756295946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,4096,0.02019999921321869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,3072,0.015778667396969266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,8192,0.03545688920550876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,3584,0.01314933349688848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,2560,0.013786666923099093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,7168,0.03134844369358487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,3584,0.016113777955373127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,3072,0.012598222328556908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,2048,0.011781333221329583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,1536,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,3072,0.01644088824590047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,2560,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,2048,0.00997155573632982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,2560,0.013586666848924426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,1536,0.009450666606426239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,2048,0.012771555946932899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,768,0.005122666557629903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,1024,0.008393777741326226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,1536,0.01069155583779017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,512,0.003563555578390757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,1024,0.00833511104186376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,768,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,16384,10240,0.028629332780838013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,768,0.007316444483068254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,128,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,512,0.00665244460105896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,256,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,64,0.0019315555691719055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,256,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,32,0.0021608887861172357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,128,0.006579555571079254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,128,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,64,0.005711110929648082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,65536,0.2034355534447564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,12288,32,0.0053484443989064955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,65536,0.14007555113898382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,51200,0.1593288845486111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,1024,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,7168,0.01984177695380317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,65536,0.1269742250442505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,51200,0.1070257756445143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,12288,256,0.0025128889828920364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,12288,0.0420702232254876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,51200,0.09986844327714707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,16384,0.04879644513130188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,10240,0.03600088755289713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,16384,0.03587733374701606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,8192,0.030118223693635728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,12288,0.0277804434299469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,12288,0.038935999075571694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,10240,0.02425244450569153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,10240,0.032807999187045626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,8192,0.02085599965519375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,8192,0.027058665951093037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,6144,0.023973332511054143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,7168,0.01921422282854716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,5120,0.020489777127901714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,7168,0.023519999451107446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,6144,0.017145777742067974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,4096,0.017472000585661996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,5120,0.016011555989583332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,6144,0.021003555920388963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,3584,0.015485333071814643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,5120,0.018580444984965853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,16384,0.05654222435421414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,3072,0.014111111561457316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,3584,0.012967111335860359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,4096,0.01648622254530589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,2560,0.012545777691735161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,12288,512,0.007130666739410824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,3072,0.011749332977665795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,3072,0.013816888961527081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,2048,0.009354666703277165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,2560,0.010652444428867765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,2560,0.01201600001917945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,2048,0.009222221871217092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,7168,0.027038221557935078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,1024,0.0058711109062035876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,1536,0.008678221868144141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,1536,0.009062222308582729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,768,0.003846222327815162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,1024,0.007675555845101674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,1024,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,512,0.0030622221529483795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,768,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,768,0.006867555694447623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,256,0.0022497777309682635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,512,0.0069368887278768755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,4096,0.013635555903116861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,512,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,256,0.006313777632183499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,3584,0.01441333360142178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,64,0.0018159999615616268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,256,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,32,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,128,0.005336888962321811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,65536,0.16733154985639784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,1536,0.007822222179836696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,64,0.005680000202523337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,32,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,10240,2048,0.011021333436171213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,51200,0.12705066468980578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,65536,0.09121511379877727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,16384,0.04315200116899279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,12288,0.03404266635576884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,65536,0.07202488846249051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,51200,0.058000889089372426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,10240,0.026702221896913316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,16384,0.022260444031821355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,16384,0.035192890299691096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,12288,0.018529777725537617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,10240,0.016167999969588388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,12288,0.02697066631582048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,7168,0.019761777586407132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,10240,0.024141333169407312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,10240,128,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,6144,0.017176000608338248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,8192,0.021381333470344543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,7168,0.013564444250530668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,10240,128,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,5120,0.014921777778201632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,6144,0.012534221841229333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,7168,0.01834488908449809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,4096,0.012726222475369772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,6144,0.016520000166363187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,5120,0.012359999948077731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,3584,0.011047110789351992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,5120,0.014793778459231058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,4096,0.01037333326207267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,3072,0.009963555468453301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,3584,0.009508444203270806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,2560,0.008616888688670265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,3584,0.011672000090281168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,3072,0.008799999952316284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,51200,0.06855199734369914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,3072,0.011439111497667102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,2048,0.007257777783605788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,2560,0.008096888661384583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,2560,0.010246222217877706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,1536,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,8192,0.022991999983787537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,2048,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,1024,0.004515555583768421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,2048,0.009519110951158736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,1536,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,768,0.0031439999325407874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,1536,0.0075839997993575195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,8192,0.014496889379289416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,512,0.0025928889711697898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,1024,0.006751110984219446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,768,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,256,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,512,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,768,0.00599911105301645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,128,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,512,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,256,0.00519466648499171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,128,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,32,0.0019528888579871918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,128,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,4096,0.012905778156386482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,64,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,65536,0.14620888233184814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,32,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,51200,0.11525866720411514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,65536,0.08183466725879245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,65536,0.06578755378723145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,51200,0.06210044357511732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,51200,0.0533884432580736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,8192,1024,0.006799110935793982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,16384,0.03353688783115811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,16384,0.021393777595625982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,12288,0.03124089042345683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,10240,0.026586666703224182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,8192,256,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,12288,0.017788444956143696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,8192,0.0225911107328203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,10240,0.01566755606068505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,12288,0.02699377801683214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,7168,0.017682666579882305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,10240,0.02299555473857456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,8192,0.014097778333557976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,6144,0.015488889482286243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,8192,0.020620443754725985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,5120,0.01349066694577535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,7168,0.018208889497650992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,6144,0.016137777103318107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,4096,0.011703111231327057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,5120,0.011225777367750803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,5120,0.014827556080288358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,16384,0.039429333474900984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,3584,0.010056888891590966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,4096,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,3072,0.009071110851234859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,4096,0.012706666356987424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,3584,0.008799110849698385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,2560,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,3584,0.011010666688283285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,3072,0.008614222208658854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,2048,0.006800000038411882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,2560,0.007320000065697565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,3072,0.010960000256697336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,1536,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,2560,0.00979555563794242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,2048,0.0070062221752272705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,1024,0.0038275557259718576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,8192,64,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,768,0.00314666661951277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,1536,0.00658044425977601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,1024,0.006259555617968242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,512,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,768,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,1024,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,256,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,768,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,512,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,128,0.0021271111650599376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,512,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,256,0.005719999886221356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,64,0.001869333287080129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,256,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,7168,32,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,128,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,64,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,65536,0.12049333254496257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,32,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,51200,0.09488444195853339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,65536,0.06302666664123535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,2048,0.008649778034951951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,16384,0.03345333205329047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,7168,1536,0.0076142218377855085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,7168,0.012869333227475485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,51200,0.05619733201132881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,12288,0.02666933337847392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,6144,0.01184177812602785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,16384,0.02108888824780782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,10240,0.02278577784697215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,16384,0.029823111163245306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,12288,0.01720355616675483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,8192,0.019659555620617337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,12288,0.024572443630960252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,10240,0.015283554792404175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,7168,128,0.005635555419656966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,7168,0.017383111847771537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,10240,0.020963556236690942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,8192,0.013551111022631327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,8192,0.0189057770702574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,7168,0.012716444830099741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,7168,0.016861332787407767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,5120,0.013569778038395775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,6144,0.01165511127975252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,6144,0.01516711049609714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,4096,0.010982222027248807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,5120,0.010384889112578498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,3584,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,65536,0.07321866353352864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,5120,0.013234666652149625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,4096,0.008718222379684448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,3072,0.008479110896587372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,4096,0.012543111211723752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,3584,0.008373333348168267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,2560,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,3584,0.010953777366214328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,3072,0.0076248885856734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,2048,0.006222222414281633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,3072,0.010838222172525195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,2560,0.007351111206743453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,1536,0.005286222116814719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,2048,0.007028444773621029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,2560,0.009030222064918941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,1024,0.00360355567600992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,1536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,2048,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,768,0.0028453332682450614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,1536,0.00702400008837382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,1024,0.006301333506902059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,512,0.002556444456179937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,6144,0.015466666883892484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,256,0.0021991110924217436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,768,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,512,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,128,0.0021128888345426982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,512,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,256,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,256,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,64,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,128,0.005700444595681296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,6144,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,128,0.004905777672926585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,65536,0.10405777560340033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,64,0.00527555536892679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,32,0.005205333232879639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,51200,0.08245511187447442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,65536,0.06709333260854085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,65536,0.06087466743257311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,16384,0.03001333276430766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,12288,0.02401955591307746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,51200,0.0507262216673957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,51200,0.04950933323966133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,16384,0.020289777053727042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,10240,0.020298666424221463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,12288,0.016568000117937725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,6144,1024,0.006334222025341458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,8192,0.017807111144065857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,12288,0.0233608881632487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,10240,0.014936889211336771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,7168,0.015814221567577787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,10240,0.01990133358372582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,6144,0.01389955480893453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,8192,0.013092444174819522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,8192,0.01787911189926995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,7168,0.01218577805492613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,5120,0.012219555675983429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,7168,0.015992000699043274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,6144,0.011367110742463006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,4096,0.00961599995692571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,6144,0.014649778604507446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,5120,0.009688888986905416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,3584,0.008627555436558193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,4096,0.008624888956546783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,5120,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,51200,0.051295999023649425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,3072,0.0075866662793689305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,4096,0.01202133297920227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,2560,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,3584,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,2048,0.005872888697518243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,3584,0.010334221853150262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,2560,0.007616889145639207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,3072,0.010067555639478896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,1536,0.004567111117972268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,2560,0.008279999925030602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,2048,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,1024,0.0032222221295038858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,2048,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,16384,0.02851111027929518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,1536,0.006627555522653792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,768,0.0028391112056043413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,6144,768,0.005967111223273807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,1536,0.006646222124497096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,1024,0.006313777632183499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,768,0.006050666587220297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,1024,0.006226666685607698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,256,0.0021991110924217436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,768,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,128,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,512,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,256,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,64,0.0018311111877361934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,128,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,128,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,32,0.0018560000591807896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,65536,0.08745777606964111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,64,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,3072,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,51200,0.06474755869971381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,65536,0.06062666575113932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,5120,512,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,51200,0.04631111025810242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,51200,0.04754044281111824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,16384,0.02461244497034285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,12288,0.019145778483814664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,512,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,16384,0.01944533321592543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,5120,256,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,10240,0.01680799987581041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,16384,0.026376888155937195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,12288,0.016132444143295288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,8192,0.014482667048772177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,12288,0.021985777550273474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,7168,0.01310755560795466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,8192,0.012755555411179861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,10240,0.018806222412321303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,6144,0.011909333368142446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,8192,0.01742755538887448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,7168,0.011368888947698804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,5120,0.010584889186753167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,7168,0.015522667103343539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,6144,0.010672888822025724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,4096,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,6144,0.013460444079505073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,5120,0.008850666383902231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,65536,0.05895377529991997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,3584,0.01069155583779017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,5120,0.013106666505336761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,3072,0.009644444617960189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,4096,0.01165066659450531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,3584,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,2560,0.008993777963850234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,3584,0.008809777597586313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,3072,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,2048,0.007165333463086023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,2560,0.007306666837798224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,1536,0.0053546664615472155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,2048,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,10240,0.014359111587206522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,1024,0.004242666479614046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,1536,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,1536,0.00676711110605134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,768,0.003560888684458203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,1024,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,5120,32,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,512,0.0028293333533737394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,768,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,768,0.005664888769388199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,256,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,512,0.0058311112225055695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,512,0.00527555536892679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,128,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,256,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,256,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,64,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,3072,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,128,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,4096,32,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,128,0.00489688871635331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,2560,0.007681777907742395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,64,0.005075555708673266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,2048,0.007583111524581909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,32,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,51200,0.05903289053175184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,65536,0.05659021933873495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,65536,0.05769244167539808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,16384,0.022327999273935955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,12288,0.017503110898865592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,51200,0.046748446093665234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,51200,0.044284443060557045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,10240,0.01517511076397366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,16384,0.025360888904995386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,12288,0.015835555063353646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,12288,0.021980444590250652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,8192,0.013134222063753339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,10240,0.014135110709402295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,4096,4096,0.008349333372380998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,7168,0.011783999701340994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,8192,0.012307555311255984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,10240,0.017457778255144756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,6144,0.010810666614108615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,8192,0.016505777835845947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,7168,0.011335111326641507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,5120,0.009865778187910715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,7168,0.014608888162506951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,6144,0.012773333324326409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,4096,0.008593777815500895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,5120,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,5120,0.011755555868148804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,3584,0.009967111051082611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,4096,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,4096,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,3072,0.008547555241319869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,3584,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,4096,1024,0.005904888941182031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,3584,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,2560,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,3072,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,3072,0.00829777783817715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,2560,0.007038222418891058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,2560,0.007886221839321984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,1536,0.0053084443012873335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,2048,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,2048,0.007079111205206976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,1024,0.003941333128346337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,1536,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,1536,0.00674222202764617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,768,0.003540444291300244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,65536,0.07599111398061116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,1024,0.006213333457708359
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,512,0.002867555452717675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,1024,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,768,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,256,0.0021564445147911706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,768,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,512,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,512,0.005294222384691238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,256,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,16384,0.01917866700225406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,64,0.0018764444523387486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,128,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,32,0.0018159999615616268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,2048,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,128,0.0047626665068997275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,65536,0.06123377879460653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,64,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,51200,0.0487937761677636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,32,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,65536,0.05459288756052653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,65536,0.05709422296947903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3584,6144,0.01036977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,16384,0.01887022289964888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,51200,0.04576444294717577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,51200,0.042704890171686806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,16384,0.01906133360332913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,12288,0.015795555379655626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,10240,0.013413333230548434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,16384,0.023680888944202002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,8192,0.011735999749766456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,12288,0.015807110402319167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,7168,0.010882666541470421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,10240,0.014096889230940076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3584,128,0.0018577778504954444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,12288,0.019309333629078336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,8192,0.011397333608733283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,6144,0.01000711073478063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,10240,0.015499555402331881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,5120,0.009176889227496253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,8192,0.015099555253982544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,7168,0.01052977806992001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3584,256,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,4096,0.008008889026112026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,6144,0.009646221995353699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,3584,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,7168,0.013398222625255585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,5120,0.008759111166000366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,3072,0.008833777573373582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,6144,0.011974222130245633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,4096,0.008174222376611497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,2560,0.00741511086622874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,5120,0.010080888867378235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,3584,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,2048,0.006315555423498154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,4096,0.008465777668688033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,3072,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,3584,0.00776977754301495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,1536,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,2560,0.006931555353932911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,1024,0.004063111212518481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,3072,0.007982222570313348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,2048,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,768,0.003896000070704354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,2560,0.007101332975758447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,1536,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,512,0.0031759999692440033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,2048,0.006643555644485686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,1024,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,1536,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,768,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,1024,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,512,0.005753777921199799
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,768,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,128,0.002191999927163124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,256,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,512,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,256,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,65536,0.05532177620463901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,128,0.004864888886610667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,51200,0.04268266757329305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,64,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,65536,0.05574666791492038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,65536,0.05208622084723579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,16384,0.017218665944205392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,51200,0.04525422387652927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,12288,0.013770666387346057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,16384,0.01888533267709944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,16384,0.021599110629823472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,10240,0.012004444168673621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,12288,0.015084443820847405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,12288,0.0174026671383116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,8192,0.010425777898894416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,10240,0.012802666260136498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,3072,256,0.002536888958679305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,7168,0.00964355551534229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,10240,0.015377778146002026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,8192,0.011058666639857821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,6144,0.008993777963850234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,7168,0.010392888552612728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,7168,0.012376888758606382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,5120,0.00794666674402025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,6144,0.009687999884287516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,3072,32,0.004865777575307422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,6144,0.010775110787815519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,4096,0.007585778004593319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,5120,0.008962666822804345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,3584,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,4096,0.008019555773999956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,5120,0.008640888664457533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,51200,0.04107111030154758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,3072,0.0074275558193524676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,4096,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,2560,0.007248889240953658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,3584,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,3584,0.007592889169851939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,3072,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,2048,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,2560,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,3072,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,1536,0.005028444445795483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,2560,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,2048,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,1024,0.003927999900446998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,1536,0.006620444357395172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,2048,0.006501333167155583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,1024,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,1536,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,8192,0.013801777528391944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,512,0.002846222163902389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,1024,0.005655999812814925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,768,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,768,0.0052888890107472735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,128,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,512,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,256,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,256,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,128,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,32,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,128,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,65536,0.047898666726218335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,64,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2560,32,0.004872888740566042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,51200,0.03664799862437778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,65536,0.05535822113355001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,65536,0.050311999188529126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,16384,0.014704000618722705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,768,0.0034355556385384668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,51200,0.04481777879926893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,12288,0.011608888705571493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,51200,0.040072888135910034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,256,0.002520888836847411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,16384,0.018228444788191054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,10240,0.010214221974213919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,16384,0.01847199930085076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,12288,0.014132445057233175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2560,512,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,8192,0.009380444056457942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,7168,0.008851555486520132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,12288,0.01519555515713162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,10240,0.012111110819710625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,6144,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,8192,0.01070755554570092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,10240,0.014473777678277759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,7168,0.01035911093155543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,6144,0.009365333451165093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,7168,0.01033333357837465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,6144,0.009029332962301042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,4096,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,5120,0.008975110948085785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,4096,0.00829777783817715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,4096,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,3072,128,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,3584,0.007315555380450354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,3072,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,3584,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2560,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,2560,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,3072,0.007703111403518253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,5120,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,2048,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,2560,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,2560,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,5120,0.008276444342401292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,1536,0.005192000004980299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,2048,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,1024,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,1536,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,3584,0.007978666987684038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,1536,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,768,0.0032355555643637977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,1024,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,1024,0.005643555687533484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,512,0.0029066666546795103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,768,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,256,0.00249422217408816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,768,0.005533333453867171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,512,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,128,0.00217333332531982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,256,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,256,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,64,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,3072,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,8192,0.012716444830099741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,128,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,2048,32,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,65536,0.03426400158140395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,128,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,51200,0.027012444204754297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,64,0.004939555707905027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,32,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,2048,2048,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,65536,0.054767999384138316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,65536,0.04827733172310723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,12288,0.009646221995353699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,51200,0.04446577694680956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,51200,0.038711110750834145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,16384,0.017222222354676988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,10240,0.0086986662613021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,12288,0.013441777891582914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,8192,0.007658667034573025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,16384,0.017610667480362784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,7168,0.00794933322403166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,12288,0.014500444134076437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,10240,0.01203555530971951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,6144,0.007263110743628607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,8192,0.010810666614108615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,10240,0.01275644451379776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,7168,0.010113777385817634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,8192,0.01038755559258991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,5120,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,2048,512,0.00563822231358952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,4096,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,7168,0.009311111436949836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,6144,0.009713778065310584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,3584,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,5120,0.008961777720186446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,6144,0.008692444198661381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,3072,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,4096,0.008303999900817871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,5120,0.008166222108734978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,2560,0.005139555368158552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,4096,0.007628444168302748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,2048,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,3584,0.007256888680987888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,3072,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,1536,0.004839999808205498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,2560,0.007272889216740926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,3072,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,1024,0.003565333369705412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,2048,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,2560,0.006552000012662675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,768,0.003152888889114062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,2048,0.006125333408514659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,16384,0.01201777739657296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,1024,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,1536,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,1024,0.0052675555149714155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,768,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,256,0.002177777803606457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,512,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,768,0.005280888742870755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,128,0.0021235555824306277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,256,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,64,0.0018728888697094387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,128,0.005542222410440445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,256,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,32,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,128,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,65536,0.023747555083698694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,64,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,32,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,3584,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,51200,0.019567999574873183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,65536,0.054139554500579834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,65536,0.046557333734300405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,16384,0.013610666824711693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,51200,0.04370933439996508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,51200,0.03736622134844462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,12288,0.011384000380833944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1536,1536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,16384,0.015944000747468736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,10240,0.009638222555319468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,16384,0.016401777664820354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1536,512,0.0026986667265494666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,8192,0.009099555512269339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,12288,0.01310488912794325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,10240,0.012064889073371887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,12288,0.013076444466908773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,10240,0.011033777561452655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,6144,0.007250666618347168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,8192,0.009662222531106737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1536,512,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,7168,0.009732444253232744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,5120,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,7168,0.009361777868535783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,4096,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,5120,0.008866666919655269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,6144,0.00868622213602066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,3584,0.00553955551650789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,5120,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,4096,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,3072,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,4096,0.007680888805124495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,3584,0.007668444679843054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,2560,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,3584,0.007259555160999298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,3072,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,2048,0.00564533347884814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,3072,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,2560,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,1536,0.004898666507667965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,2560,0.00630044440428416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,2048,0.006917333437336816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,2048,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,7168,0.008392888638708327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,1536,0.006130666782458623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,768,0.0032257777121331957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,1536,0.0058497778243488735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,1024,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,512,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,768,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,768,0.005290666802061929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,6144,0.009314667019579146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,256,0.002504000026318762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,512,0.00554044461912579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,128,0.0021235555824306277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,512,0.005076444397370021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,256,0.005558222118351195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,64,0.0018159999615616268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,256,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,32,0.0018613334331247543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,128,0.004856889032655292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,64,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,32,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,51200,0.015747555428081088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,65536,0.045781334241231285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,65536,0.05397599935531616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,1024,1024,0.003556444413132138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,16384,0.013037333057986366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,8192,0.01036266651418474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,51200,0.04313777883847555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,12288,0.009962666365835402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,16384,0.015805333852767944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,51200,0.0369777778784434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,10240,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,16384,0.014481777946154276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,8192,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,12288,0.012062222593360476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,10240,0.01109155515829722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,7168,0.007095110913117726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,8192,0.01036088913679123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,8192,0.00999466660949919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,6144,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,1024,128,0.005404444618357553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,7168,0.009768000079525841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,5120,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,6144,0.009224889179070791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,6144,0.008633777499198914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,4096,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,5120,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,5120,0.007981333467695449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,3584,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,4096,0.007677333222495184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,4096,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,3072,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,3584,0.00738844441043006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,3584,0.00721955547730128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,12288,0.013062222136391534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,2560,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,10240,0.011759999725553723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,3072,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,2048,0.004265777766704559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,3072,0.0069857777820693115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,2560,0.0069066666894488865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,1536,0.004519111166397731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,2560,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,65536,0.018888889087571036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,2048,0.005987555616431766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,7168,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,1536,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,768,0.0031662223239739737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,1536,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,1024,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,512,0.002542222125662698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,1024,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,768,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,256,0.002193777718477779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,768,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,512,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,128,0.0019617778145604664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,512,0.004987555659479565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,256,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,256,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,1024,1024,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,32,0.001810666587617662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,65536,0.026836444934209187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,128,0.0047360000511010485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,64,0.004863111095296012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,51200,0.02223644488387638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,65536,0.051382223765055336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,65536,0.0451546675629086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,16384,0.00906755526860555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,51200,0.03834755553139581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,768,1024,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,12288,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,16384,0.01549600064754486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,16384,0.014448889427714877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,10240,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,12288,0.012930666406949362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,12288,0.01238844460911221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,8192,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,10240,0.011735999749766456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,10240,0.01073688848151101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,7168,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,128,0.005283555636803309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,8192,0.010021333065297868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,6144,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,7168,0.009716444545321995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,7168,0.009332444104883406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,768,32,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,5120,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,6144,0.009320889082219865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,6144,0.008442666795518663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,768,2048,0.00646311127477222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,3584,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,5120,0.008610666626029545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,4096,0.008163555628723567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,5120,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,51200,0.03633600142267015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,3072,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,3584,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,4096,0.007095110913117726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,3584,0.007224000162548489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,3072,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,2048,0.004644444419278039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,2560,0.00692622239391009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,3072,0.006721777634488211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,2560,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,2048,0.006440889090299606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,1024,0.0029582221888833572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,2048,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,1536,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,768,0.0028213332924577924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,1536,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,8192,0.010377777947319878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,512,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,1024,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,768,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,256,0.002230222233467632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,768,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,512,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,128,0.0019404445257451797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,4096,0.005909333212508096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,256,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,64,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,256,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,32,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,128,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,65536,0.016680889659457736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,128,0.00461777796347936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,2560,0.005192000004980299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,51200,0.01371911085314221
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,64,0.004563555535342959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,32,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,65536,0.04622755448023478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,512,1536,0.0035582222044467926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,16384,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,51200,0.037343111303117536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,12288,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,51200,0.03461866577466329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,16384,0.015491555134455362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,10240,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,16384,0.014204444156752693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,12288,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,512,1024,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,8192,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,12288,0.011864888999197217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,10240,0.011779555843936073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,8192,0.01068800025516086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,10240,0.010904889139864178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,6144,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,8192,0.009650666680600908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,512,512,0.004903111192915174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,5120,0.004574222283230888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,7168,0.008986666798591614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,6144,0.009131555755933126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,4096,0.004233777936961916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,6144,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,5120,0.008632000121805403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,3584,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,5120,0.007631111476156447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,4096,0.007687999970383114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,3072,0.004547555413511065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,4096,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,3584,0.007544000115659501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,65536,0.042484445704354175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,2560,0.004577777865860197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,3584,0.007118222614129384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,2048,0.004536889079544279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,3072,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,2560,0.006992888947327931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,1536,0.0035333335399627686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,2560,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,2048,0.006342222293217977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,2048,0.005894222193294101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,7168,0.005201777650250329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,1536,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,1536,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,1024,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,7168,0.010022222167915767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,1024,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,768,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,768,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,256,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,512,0.0053048887186580235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,512,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,128,0.0018088887963030075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,256,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,256,0.004588444613748127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,64,0.0018133332745896445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,32,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,128,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,256,3072,0.007253333098358578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,128,0.004594666676388847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,64,0.004588444613748127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,65536,0.04584533307287428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,51200,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,1024,0.00286311118139161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,65536,0.04154400030771891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,16384,0.00543200017677413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,51200,0.037003556887308754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,12288,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,16384,0.015772443678643968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,51200,0.033555554019080266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,10240,0.005374222166008419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,512,0.002467555511328909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,16384,0.014133334159851074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,8192,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,12288,0.0127688886390792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,10240,0.012070222033394707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,12288,0.011791999969217511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,7168,0.005193777796294955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,8192,0.01056177748574151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,6144,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,8192,0.009638222555319468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,7168,0.010038221875826517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,5120,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,7168,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,6144,0.009217778013812171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,4096,0.004197333421972063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,65536,0.011353777514563667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,6144,0.008375110725561777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,5120,0.008563555777072906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,3584,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,5120,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,3072,0.004573333180612988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,256,32,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,3584,0.007344000041484833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,4096,0.007255111303594377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,2560,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,3584,0.006695111178689533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,3072,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,256,768,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,3072,0.0069671107663048645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,2560,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,2048,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,1536,0.0035582222044467926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,2048,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,1024,0.002871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,1536,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,10240,0.010802666346232096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,1536,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,768,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,1024,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,512,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,1024,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,768,0.0058346668051348785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,256,0.001960888918903139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,768,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,512,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,128,0.0018648888087934917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,512,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,256,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,4096,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,64,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,256,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,32,0.0018133332745896445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,65536,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,128,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,51200,0.006608888920810487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,2560,0.006692444284756978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,64,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,128,32,0.004537777768241035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,16384,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,12288,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,10240,0.004538666870858935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,65536,0.04176000091764662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,8192,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,51200,0.03391999999682108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,16384,0.014127110441525778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,7168,0.004943999979231092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,12288,0.011896000140243106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,6144,0.004550222307443619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,5120,0.004556444370084339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,10240,0.010695999695195092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,4096,0.0041982221106688184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,8192,0.009717333647939894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,3584,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,7168,0.008968888885445064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,6144,0.008361777497662438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,5120,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,2560,0.004560889055331548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,4096,0.006973333656787872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,2048,0.003914666672547658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,3584,0.007035555938879649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,1536,0.0035075553589397003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,3072,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,5,128,128,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,1024,0.0028515555378463534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,2560,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,2048,0.0059075554211934405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,128,2048,0.004217777815130022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,512,0.0022213332768943575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,256,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,1536,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,128,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,1024,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,64,0.0018240000224775737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,768,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,32,0.0018275556051068837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,65536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,512,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,256,0.004549333204825719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,51200,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,16384,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,64,0.004519111166397731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,12288,0.004845333182149463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,65536,0.041951109965642296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,10240,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,51200,0.033237334754731916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,16384,0.014136888914638095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,8192,0.003957333250178231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,7168,0.004288889053795072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,12288,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,6144,0.004179555508825514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,10240,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,5120,0.004549333204825719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,8192,0.009364444348547194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,4096,0.0038524443904558816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,7168,0.009036444127559662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,768,0.002519999941190084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,6144,0.00810933361450831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,3072,0.0038444445365005066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,5120,0.0079297779334916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,2560,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,4096,0.0069466663731469055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,2048,0.003908444609906939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,3584,0.0069431112044387394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,1536,0.003543111185232798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,3072,0.006321777900060018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,1024,0.002817777709828483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,2560,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,768,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,2048,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,32,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,1536,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,1024,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,256,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,64,3072,0.0038524443904558816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,768,0.005195555587609609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,128,0.0017804443422291016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,64,0.001503999965886275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,512,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,32,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,256,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,128,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,64,0.004521777646409141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,3584,0.004108444270160464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,32,32,0.004705777598751916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,16384,0.33458399772644043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,5,64,128,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,5,32,512,0.0022400000856982335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,51200,1.075301382276747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,16384,0.2286977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,16384,0.13541066646575928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,12288,0.2541573312547472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,12288,0.15597689151763916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,51200,0.44147200054592556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,10240,0.2126071055730184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,51200,0.48563554551866317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,10240,0.15767378277248806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,12288,0.10199822319878472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,8192,0.16438844468858507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,10240,0.08738578028149074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,16384,0.17321511109670004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,8192,0.11865422460767959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,12288,0.13098755147722033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,8192,0.07089689042833117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,7168,0.1034346686469184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,6144,0.12472089131673177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,7168,0.07159466875924005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,10240,0.11189156108432347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,5120,0.10144088665644328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,6144,0.08883111344443427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,6144,0.06977333625157674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,8192,0.0937439997990926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,7168,0.08249688810772367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,4096,0.08382133642832439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,5120,0.08282044198777941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,5120,0.05747555361853706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,6144,0.07237422466278076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,3584,0.07210044728385077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,7168,0.14244000116984049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,4096,0.06799555487102933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,5120,0.06298577785491943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,3072,0.0624177787039015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,4096,0.03851200143496195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,4096,0.05264888869391548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,3584,0.034747554196251765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,3584,0.0543084442615509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,2560,0.052367998494042285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,3072,0.050436443752712674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,3072,0.03064889046880934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,2048,0.04288888971010844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,2560,0.04766577813360426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,2560,0.02878755662176344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,3072,0.04292177822854784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,1536,0.03267555435498556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,2560,0.03830133212937249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,2048,0.037893331713146634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,2048,0.028207110034094915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,1024,0.02274488906065623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,1536,0.023408000667889912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,2048,0.03372800019052293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,1536,0.028044445647133723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,768,0.017775999175177682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,1536,0.028563555743959215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,1024,0.020822222034136455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,512,0.012702222499582501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,1024,0.02273333403799269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,768,0.017180444465743173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,768,0.015184000134468079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,1024,0.01794666714138455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,256,0.007589333587222629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,512,0.01275466630856196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,512,0.013064000341627332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,768,0.020278222031063504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,128,0.005036444299750858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,512,0.017423111531469557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,3584,0.04831555485725403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,256,0.010370666782061258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,64,0.004223111189073986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,65536,32,0.0041626666982968645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,51200,0.5027724372016059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,128,0.01035644445154402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,128,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,256,0.01309511148267322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,128,0.010437332921557956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,64,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,32,0.00908622228437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,65536,0.5116524696350098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,51200,0.3882435427771674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,51200,0.7884391148885092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,65536,1.0081511603461373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,65536,256,0.010320000350475311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,16384,0.2554871241251628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,16384,0.1525582207573785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,65536,0.43952178955078125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,12288,0.18991022639804414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,51200,0.3495306703779433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,12288,0.1521893342336019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,10240,0.15700444910261366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,12288,0.08319822284910414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,16384,0.1383848852581448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,10240,0.1250657770368788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,16384,0.11210489273071289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,8192,0.1289493375354343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,10240,0.06853422191407946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,12288,0.10667200220955743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,7168,0.11400266488393147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,8192,0.05676266882154676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,10240,0.09159466955396865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,8192,0.09432533052232529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,7168,0.08528355757395427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,6144,0.09911200073030259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,8192,0.07726222276687622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,6144,0.0742115577061971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,7168,0.051106668180889554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,7168,0.06715644730461968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,5120,0.0888142254617479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,6144,0.04678488771120707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,5120,0.0687733358807034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,6144,0.06011822488572863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,5120,0.03828888800409105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,4096,0.032413333654403687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,3584,0.06272000074386597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,4096,0.05484355489412943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,51200,0.4223991235097249
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,4096,0.04454311066203647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,3072,0.054663111766179405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,3584,0.029454221328099568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,3584,0.04401155643992954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,2560,0.045399109522501625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,3584,0.041747556792365186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,3072,0.042809777789645724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,3072,0.026199999782774184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,2048,0.03659644391801622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,3072,0.037964443365732826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,2560,0.02402044501569536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,2560,0.037003556887308754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,1536,0.02846577763557434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,2560,0.033024887243906655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,2048,0.03100799851947361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,2048,0.02308533257908291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,4096,0.07125777668423124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,1024,0.020230222079488967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,65536,0.5135244263543023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,1536,0.019903111788961623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,2048,0.029937777254316542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,1536,0.023265777362717524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,768,0.014893333117167154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,5120,0.051912890540228955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,1024,0.017362667454613578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,512,0.010405333505736457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,1024,0.01575288838810391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,768,0.013582222163677216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,256,0.006263111200597551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,768,0.014797333214018079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,512,0.011026666396194033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,512,0.011387555963463254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,512,0.015427556302812366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,128,0.0031946665710873077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,256,0.008396444221337637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,256,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,64,0.0025439999169773525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,256,0.012040889097584618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,51200,32,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,51200,128,0.009328000247478485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,128,0.007596444752481248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,128,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,64,0.008350222474998897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,51200,32,0.00833511104186376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,65536,0.31624799304538304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,51200,0.2475395467546251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,65536,0.17145778073204887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,1536,0.024801777468787298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,65536,0.1520035531785753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,16384,0.08380533589257134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,51200,0.13075378206041124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,768,0.017779555585649278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,12288,0.06452533271577623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,51200,0.12320000595516628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,16384,0.042240000433391996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,12288,0.04811911119355095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,12288,0.032800889677471586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,16384,0.05577244361241659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,10240,0.05576444334454007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,12288,0.044623110029432506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,8192,0.04193688763512505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,10240,0.04312888781229655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,10240,0.030992888742023047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,10240,0.038782222403420344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,7168,0.037131554550594754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,51200,1024,0.0199297782447603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,8192,0.03750844465361701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,8192,0.03224888775083754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,65536,0.17381244235568574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,6144,0.03187111020088196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,8192,0.03360266817940606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,7168,0.03475111060672336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,7168,0.026535999443795946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,5120,0.02752444479200575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,7168,0.031488888793521456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,16384,0.06424888637330797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,6144,0.019902222686343722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,6144,0.02869422237078349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,4096,0.023211555348502264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,6144,0.02828177809715271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,5120,0.017417778571446735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,5120,0.025570667452282373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,3584,0.019712888532214694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,4096,0.021962665849261816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,5120,0.025969778498013813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,3072,0.01717688971095615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,3584,0.01440622243616316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,3584,0.01882844501071506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,3072,0.013321778012646569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,3072,0.018012444178263348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,2560,0.015023999743991427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,3072,0.019179556104871962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,51200,0.1445351176791721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,2048,0.012336889074908363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,2560,0.012703999876976013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,2560,0.01479644411140018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,2560,0.017669333351982962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,1536,0.010213333699438307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,2048,0.014114666316244336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,2048,0.01276266657643848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,2048,0.016181333197487723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,1536,0.011032889286677042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,1536,0.011657777759763928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,4096,0.015448888142903646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,4096,0.022620444496472675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,768,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,1024,0.009002666506502364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,1024,0.009311999711725447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,1024,0.013420444395807056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,3584,0.021032000581423443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,768,0.008292444050312042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,768,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,768,0.01236355553070704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,256,0.002526222210791376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,512,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,512,0.011063111325105032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,256,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,256,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,128,0.0021893332401911416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,256,0.011025778121418424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,64,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,128,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,1024,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,32,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,128,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,128,0.009008888569143083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,1536,0.015319999721315173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,64,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,16384,32,0.0057884446448749965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,65536,0.24373510148790148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,16384,512,0.0041751112374994485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,51200,0.18774844540490043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,16384,0.064355558819241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,65536,0.13406933678521052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,512,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,51200,0.11690755685170491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,16384,0.03769244419203864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,16384,0.05448089043299357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,12288,0.05006577902370029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,16384,0.044906665881474815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,10240,0.04223555657598707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,12288,0.029975109630160864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,12288,0.04445866743723551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,12288,0.03542755709754096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,10240,0.025682666235499915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,8192,0.036679999695883855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,51200,0.10564888848198785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,10240,0.03705866800414191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,65536,0.1384915510813395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,7168,0.033013333876927696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,10240,0.032031112247043185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,65536,0.14761333995395237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,8192,0.02168888847033183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,6144,0.02878755662176344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,8192,0.03288177649180094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,7168,0.023248000277413264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,51200,0.11132178041670059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,8192,0.02863466739654541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,6144,0.017895999881956313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,6144,0.0258933338854048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,4096,0.020959999826219346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,6144,0.023736000061035156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,5120,0.016015999846988253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,5120,0.023033777872721355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,3584,0.0188417782386144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,5120,0.02160444524553087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,4096,0.01421866648726993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,4096,0.0203146669599745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,4096,0.019370666808552213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,3072,0.01626666718059116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,3584,0.01677600045998891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,3584,0.018533334136009216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,3072,0.016699555847379897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,3072,0.012264889147546558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,2560,0.014320888453059726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,3072,0.0173688895172543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,5120,0.02491377790768941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,2560,0.011618666350841522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,7168,0.029194666279686823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,2048,0.012460444536474017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,2560,0.01498933302031623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,7168,0.026561778452661302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,1536,0.008264000217119852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,2560,0.016183111402723525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,2048,0.011081777513027191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,1024,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,2048,0.01294311136007309
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,1536,0.011042666931947073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,1536,0.009866666462686326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,768,0.005127111242877113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,1536,0.014175110393100314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,1024,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,1024,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,512,0.003173333282272021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,3584,0.013199999928474426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,1024,0.0118568887313207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,768,0.007696888513035244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,256,0.002699555622206794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,768,0.01069155583779017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,512,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,512,0.006559111177921295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,128,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,256,0.00638488887084855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,512,0.010378667049937779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,256,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,64,0.002064888882968161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,256,0.010132444401582083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,128,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,12288,32,0.0019155556543005838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,128,0.005251555393139522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,128,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,64,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,65536,0.2018453280131022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,32,0.005601777798599667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,65536,0.14135732915666369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,51200,0.15905243820614284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,65536,0.12850756115383571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,12288,768,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,16384,0.05621955792109171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,51200,0.10679733090930515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,51200,0.10300088591045803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,12288,0.043149332205454506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,16384,0.047129776742723256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,16384,0.041033777925703264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,10240,0.03706044289800856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,12288,0.02773688899146186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,12288,0.038482666015625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,65536,0.1168702178531223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,12288,0.03273333443535699
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,51200,0.09344089031219482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,8192,0.030517333083682593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,10240,0.02466933263672723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,10240,0.032848887973361544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,7168,0.027436445156733196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,8192,0.02485688858562046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,8192,0.02723466687732273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,8192,0.02641244398223029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,6144,0.02462399999300639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,7168,0.02108977735042572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,7168,0.024715556038750544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,5120,0.020851555797788832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,7168,0.023909333679411147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,6144,0.017349332571029663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,6144,0.020972443951500788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,4096,0.017720000611411203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,6144,0.022638221581776936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,16384,0.03527999917666117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,5120,0.019186667270130582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,5120,0.015478221906556023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,3584,0.016102222932709586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,5120,0.01995377739270528
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,4096,0.013791999883121915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,3072,0.013958222336239286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,3584,0.013009777499569787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,4096,0.016607999801635742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,3584,0.014141332772043018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,4096,0.018798222144444782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,2560,0.01311466677321328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,3584,0.018213333355055917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,3072,0.011535111400816174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,2048,0.010278222461541494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,3072,0.013735999663670858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,2048,0.015916445189052157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,2560,0.010844444235165914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,3072,0.01647644407219357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,10240,0.0291920006275177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,2560,0.01202488856183158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,2048,0.010257778068383535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,2048,0.011072888970375061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,1024,0.005763555566469829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,2048,0.015021334091822306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,1536,0.009080000221729279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,1536,0.013457777599493662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,1024,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,768,0.0041999999019834726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,1024,0.007443555527263218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,1024,0.01109155515829722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,512,0.003019555575317807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,768,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,768,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,768,0.011413333316644033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,256,0.002440888848569658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,512,0.006953777538405524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,512,0.01035733355416192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,256,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,1536,0.007855999800893996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,128,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,256,0.006312888943486744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,256,0.010713777608341642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,2560,0.015357333752844067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,128,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,32,0.001981333312061098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,128,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,128,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,64,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,65536,0.16752978165944418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,32,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,65536,0.09047733412848578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,51200,0.12690488497416177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,65536,0.06990489032533434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,512,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,16384,0.04514044523239136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,51200,0.06868444548712836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,51200,0.05817777580685086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,10240,64,0.0018604445374674266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,16384,0.03564711080657111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,65536,0.11001155773798625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,16384,0.02238311039076911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,10240,1536,0.009656888743241629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,10240,0.028594666057162817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,51200,0.08892799748314752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,16384,0.03811733259095086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,12288,0.01847199930085076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,8192,0.023212444451120164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,12288,0.028578665521409776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,12288,0.031184001101387873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,10240,0.01648355523745219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,7168,0.019676445258988273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,10240,0.025758221745491028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,8192,0.014468444718254937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,10240,0.0281333327293396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,6144,0.01744888888465034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,7168,0.013569778038395775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,8192,0.024507555696699355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,7168,0.018623110320832994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,5120,0.014883556299739413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,7168,0.0229751103454166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,6144,0.012427555190192329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,4096,0.01295466638273663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,6144,0.016548444827397663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,6144,0.02103288968404134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,5120,0.011557333171367645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,5120,0.015043555034531487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,3584,0.011049778097205691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,5120,0.019471999671724107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,4096,0.013633777697881063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,12288,0.035036444664001465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,3072,0.00999111102686988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,4096,0.017825777331988018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,3584,0.009667555491129557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,3584,0.011693333586057028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,2560,0.00869955536392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,3584,0.01702577703528934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,3072,0.008961777720186446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,2048,0.007455999652544658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,3072,0.0116462219092581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,3072,0.01614755557643043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,2560,0.008347555167145198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,1536,0.006317333214812809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,2560,0.010093332992659675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,2560,0.015214223000738354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,2048,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,8192,0.02163999941613939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,1024,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,1536,0.007638221813572778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,1536,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,768,0.00318755561278926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,1536,0.01277955538696713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,1024,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,1024,0.011078221930397881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,512,0.0028088889602157804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,768,0.006272000157170826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,768,0.006271111054552927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,4096,0.010556444525718689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,256,0.002463111033042272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,768,0.011376000112957425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,512,0.006042666733264923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,512,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,512,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,256,0.005208000126812193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,128,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,256,0.005607111172543631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,256,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,128,0.005672889037264719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,64,0.001828444500764211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,8192,32,0.0018746666610240936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,128,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,128,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,2048,0.014134221606784396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,64,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,65536,0.14697333176930746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,32,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,1024,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,51200,0.11536178323957656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,65536,0.0818435549736023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,65536,0.0654568870862325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,16384,0.040765331851111516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,51200,0.06234221988254123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,51200,0.05330666568544176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,16384,0.02156177825397915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,12288,0.031930665175120033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,16384,0.033719112475713096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,16384,0.03715733355946011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,10240,0.02719466719362471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,12288,0.017812444104088675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,51200,0.088919997215271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,12288,0.027128888501061335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,8192,0.023358222511079576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,12288,0.03074755602412754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,8192,2048,0.009305777649084726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,10240,0.015836444165971544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,7168,0.017895999881956313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,10240,0.023965333898862202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,10240,0.02751111156410641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,6144,0.01592177814907498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,8192,0.0203457772731781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,8192,0.024360888534122046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,7168,0.01311911145846049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,5120,0.013674666484196981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,7168,0.018560000591807894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,7168,0.02233244478702545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,6144,0.01204177737236023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,4096,0.011980444192886353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,6144,0.02070577773782942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,5120,0.010992888775136737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,5120,0.014717333846622042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,3584,0.010132444401582083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,5120,0.019612444771660697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,4096,0.013095999757448832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,4096,0.009520889156394536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,3072,0.009217778013812171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,65536,0.11036088731553818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,4096,0.0177208897140291
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,3584,0.008768888811270395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,2560,0.00811644477976693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,3584,0.011404444773991903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,3584,0.01745244363943736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,3072,0.00832622249921163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,3072,0.011117333339320289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,2048,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,2560,0.008016889293988546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,8192,0.01402666668097178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,3072,0.01648622254530589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,1536,0.005716444303592046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,2560,0.010228444304731157
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,2048,0.007644444704055786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,2048,0.00871022211180793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,1024,0.003925333420435588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,2048,0.014205333259370593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,1536,0.007924444145626491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,6144,0.01754666699303521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,768,0.0032115555885765287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,1536,0.012060444388124677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,1024,0.006315555423498154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,512,0.0028168888141711554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,1024,0.00646311127477222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,768,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,1024,0.011767999993430244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,256,0.0023440000497632557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,768,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,768,0.011029333704047732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,512,0.005903999838564131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,512,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,128,0.002157333410448498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,512,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,256,0.005879111174080107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,256,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,7168,32,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,128,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,128,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,128,0.0083137775460879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,65536,0.12180088626013862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,64,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,7168,32,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,51200,0.09567022323608398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,2560,0.015846222639083862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,65536,0.07499555746714275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,1536,0.006679111056857639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,16384,0.03600088755289713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,65536,0.06338755289713542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,51200,0.056256890296936035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,51200,0.05117333266470167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,12288,0.02773244513405694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,16384,0.020861332615216572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,65536,0.10961600144704182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,16384,0.03737511237462362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,10240,0.023855111665195886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,51200,0.08861333131790161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,12288,0.024695111645592585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,12288,0.01715555621518029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,256,0.009879111415810054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,8192,0.020311999652120803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,12288,0.030784000953038532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,10240,0.01519377695189582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,10240,0.02202400068442027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,7168,0.017952000101407368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,10240,0.027635556128289964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,8192,0.013561777770519257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,6144,0.016016888949606154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,8192,0.023765333824687537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,7168,0.01759466694460975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,7168,0.022251556317011516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,5120,0.013756444056828817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,6144,0.011731555892361535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,6144,0.020614221692085266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,5120,0.013942221800486246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,16384,0.030646221505271062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,4096,0.010972444381978778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,5120,0.010437332921557956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,5120,0.01890222231547038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,3584,0.009322666459613377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,4096,0.009149333669079674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,4096,0.012639111114872826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,4096,0.01740711099571652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,3072,0.008212444682916006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,3584,0.010994666980372535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,3584,0.008661333057615492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,3584,0.016534222496880423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,2560,0.007148444652557373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,3072,0.007609777980380588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,3072,0.01071911139620675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,3072,0.015828443898095023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,2048,0.006167999986145232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,2560,0.0076248885856734375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,2560,0.009319110876984065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,6144,0.01555199921131134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,1536,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,2560,0.014759999182489185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,2048,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,2048,0.008764444126023186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,1024,0.003903999924659729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,1536,0.006679111056857639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,1536,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,768,0.003104000041882197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,1536,0.011676444775528379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,1024,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,1024,0.011028444601429833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,512,0.002481777841846148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,768,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,768,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,8192,0.019897777173254225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,512,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,256,0.0021671110557185277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,7168,0.012810666528013019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,512,0.010664888554149203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,128,0.0019004444281260173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,256,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,256,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,256,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,64,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,128,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,128,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,2048,0.013259555730554791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,128,0.008274666965007782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,64,0.005203555441564984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,65536,0.10592177841398452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,6144,32,0.0052062223354975385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,51200,0.08243378003438313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,768,0.00598933340774642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,65536,0.06698399782180786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,65536,0.060990220970577665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,16384,0.031369778845045306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,512,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,51200,0.04936088787184822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,65536,0.10911466677983601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,16384,0.028463999430338543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,12288,0.025349333882331848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,16384,0.019944889677895438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,51200,0.08835289213392471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,10240,0.02125955621401469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,16384,0.03638755612903171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,12288,0.016582222448454965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,6144,32,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,8192,0.01889244384235806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,12288,0.0236826671494378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,12288,0.030124445756276447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,10240,0.014813333749771118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,10240,0.02129777769247691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,7168,0.016176889340082806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,1024,0.006330666856633292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,10240,0.02648888859483931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,8192,0.013119999733236102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,8192,0.017519111434618633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,6144,0.014517333772447375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,7168,0.0123511114054256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,8192,0.02351911034848955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,5120,0.012343111137549082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,7168,0.01717511150572035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,7168,0.022335110439194575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,6144,0.011360000405046674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,6144,0.014704000618722705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,4096,0.009666666388511658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,6144,0.020265777905782063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,51200,0.050738665792677135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,3584,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,5120,0.009858667022652095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,5120,0.014144889182514615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,4096,0.008412444757090675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,4096,0.011796444654464722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,3072,0.007661333514584436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,3584,0.00793599999613232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,4096,0.01725955473052131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,3584,0.010466666685210334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,2560,0.006683555742104848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,3584,0.016751110553741455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,3072,0.009645333720578088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,3072,0.015112888481881885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,2048,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,2560,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,2560,0.008013333711359236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,1536,0.004447111239035924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,2560,0.014085332552591959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,2048,0.00699733363257514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,2048,0.007835555407736037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,1024,0.0031742221779293488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,1536,0.006956444846259222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,2048,0.013048888908492194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,1536,0.007251555720965068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,768,0.0028195555011431375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,1536,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,1024,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,1024,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,512,0.00249777775671747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,768,0.006160000132189856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,768,0.005924444645643234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,768,0.010324444207880232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,256,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,512,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,512,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,3072,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,512,0.010685332947307162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,128,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,256,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,256,0.005193777796294955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,64,0.001820444439848264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,256,0.00979911122057173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,128,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,5120,32,0.0018444444156355327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,128,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,128,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,65536,0.0898817777633667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,64,0.00508622204264005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,51200,0.06830222076839872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,5120,32,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,65536,0.0587386687596639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,16384,0.026008889079093933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,51200,0.04629244407018026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,51200,0.04736977815628052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,1024,0.011416888899273343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,5120,0.018609777092933655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,12288,0.019651555352740817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,65536,0.10925510856840347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,16384,0.019557333654827543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,16384,0.02647022240691715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,10240,0.016864889197879367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,12288,0.015929778416951496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,16384,0.036074668169021606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,8192,0.01496266656451755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,12288,0.022145777940750122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,12288,0.0295448899269104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,10240,0.01441511180665758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,10240,0.020266667008399963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,7168,0.012917333179050021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,10240,0.026583110292752583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,8192,0.012772444221708508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,8192,0.017429333594110277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,6144,0.012096888489193387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,7168,0.011686222420798408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,8192,0.023354666100607976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,7168,0.015832000308566622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,5120,0.010685332947307162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,7168,0.021554667088720534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,6144,0.010067555639478896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,6144,0.014474666780895658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,4096,0.009415999882751042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,65536,0.06035288837220934
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,6144,0.020579554968410067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,5120,0.009363555245929295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,5120,0.013377777404255338
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,3584,0.01071466671095954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,5120,0.018583999739752877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,4096,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,4096,0.011721777419249216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,3072,0.009119999905427298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,4096,0.017352888981501263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,3584,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,51200,0.08814577923880683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,2560,0.008360889222886827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,3584,0.01551822159025404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,3072,0.009022222624884712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,2048,0.007225777539942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,3072,0.01496355566713545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,2560,0.007161777880456712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,2560,0.007285333342022366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,1536,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,2560,0.013277333643701343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,2048,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,2048,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,1024,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,2048,0.013555555707878537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,1536,0.006575111299753189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,1536,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,768,0.003802666647566689
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,1536,0.011761777930789523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,1024,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,1024,0.0058782220714622075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,512,0.0031795555518733132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,1024,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,768,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,768,0.01071911139620675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,256,0.0022195554855797025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,512,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,512,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,3584,0.009296889106432596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,128,0.0018853334089120228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,512,0.010699555277824402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,256,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,256,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,64,0.0019511110666725372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,256,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,128,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,4096,32,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,128,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,128,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,65536,0.07758133278952704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,64,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,4096,32,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,51200,0.06241866615083483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,65536,0.057040890057881675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,65536,0.05784444676505195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,16384,0.023727999793158636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,768,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,65536,0.1090008881356981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,12288,0.018260444204012554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,16384,0.024830222129821777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,16384,0.03592711024814182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,3072,0.007632888853549957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,10240,0.01572711103492313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,12288,0.020915556285116408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,12288,0.015782222151756287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,12288,0.02944799926545885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,8192,0.013316444224781461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,10240,0.01408266690042284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,10240,0.018789332773950364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,10240,0.02648266653219859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,8192,0.015507555670208402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,8192,0.01219466659757826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,8192,0.02331377731429206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,51200,0.04424800144301521
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,51200,0.04660266637802124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,6144,0.011010666688283285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,7168,0.014152889450391134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,7168,0.02201066745652093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,16384,0.01923022170861562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,6144,0.009352888498041365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,6144,0.013467555244763693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,6144,0.020271110865804885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,4096,0.008309333688682979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,5120,0.009000889129108852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,5120,0.012448888685968188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,5120,0.01866222255759769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,3584,0.00958577791849772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,4096,0.01068800025516086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,4096,0.01702044407526652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,7168,0.01182755579551061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,3584,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,3072,0.008645333349704742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,3584,0.015208888385030957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,7168,0.011362666885058085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,3072,0.007630222373538547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,3072,0.00793422261873881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,3072,0.014488000008794995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,51200,0.08823821942011516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,2048,0.006437333507670297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,2560,0.0075911109646161394
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,2560,0.007426666716734569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,2560,0.013384888569513956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,2048,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,1536,0.00498844434817632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,2048,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,4096,0.00832088871134652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,2048,0.012785777449607849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,1024,0.004182222402758068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,1536,0.006612444503439798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,1536,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,1536,0.011709333293967776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,3584,0.007984889050324758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,1024,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,1024,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,1024,0.011735111474990845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,512,0.0032311110860771606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,768,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,2560,0.00759733302725686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,768,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,256,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,5120,0.009953777823183272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,768,0.01038577738735411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,512,0.0058453331391016645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,128,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,256,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,512,0.010729778144094678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,256,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,64,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,256,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,32,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,128,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,128,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,128,0.00832177781396442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,65536,0.0675413343641493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,64,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,32,0.004928888960017098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,51200,0.053202665514416166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,65536,0.05478488736682468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,65536,0.05699111355675591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,16384,0.01993777851263682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,51200,0.045722666713926524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,51200,0.042529778348075024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,65536,0.1092275579770406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,12288,0.015976000163290236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,51200,0.08767022026909722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,16384,0.02221244408024682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,16384,0.01901511185699039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,10240,0.014116444521480136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3584,512,0.0053280000057485365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,16384,0.035624000761244036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,12288,0.01979822251531813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,8192,0.012034667034943899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,12288,0.029149333635965984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,10240,0.016970665918456185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,7168,0.011256888508796692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,10240,0.026520889666345384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,8192,0.011409777734014722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,8192,0.015134221977657743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,6144,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,8192,0.023175110419591267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,7168,0.01036444471942054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,7168,0.013403555585278405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,5120,0.009312888814343346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3584,768,0.0034968890249729156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,6144,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,6144,0.012432888978057437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,4096,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,6144,0.02001066671477424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,5120,0.008955555657545725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,5120,0.010950222611427307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,3584,0.0074480002125104266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,5120,0.018464000688658822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,4096,0.008270222279760573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,4096,0.009357333183288574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,3072,0.008377778033415476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,12288,0.015855110353893705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,3584,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,4096,0.016172443826993305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,3584,0.007883555359310573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,2560,0.00721955547730128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,3584,0.015479111009173922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,10240,0.013760888742076026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,3072,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,2048,0.006133333262470033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,3072,0.007643555601437886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,2560,0.007236444287829929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,3072,0.014517333772447375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,1536,0.005946666830115848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,2048,0.006663110935025745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,2560,0.013196444345845116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,1024,0.004229333251714706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,2048,0.012504888905419244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,1536,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,1536,0.006255110932721033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,768,0.00350222239891688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,7168,0.022138666775491502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,1536,0.01202400028705597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,1024,0.006043555421961679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,1024,0.005628444254398346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,512,0.0032035555276605818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,768,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,1024,0.010787555740939246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,768,0.005489777773618698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,256,0.0025475554996066624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,768,0.011053333679835001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,512,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,512,0.005568000177542369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,128,0.0022640000614855024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,512,0.01038044442733129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,256,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,256,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,64,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,256,0.0103502223889033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,128,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,3072,32,0.0018133332745896445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,2560,0.006994666324721442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,65536,0.057208889060550265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,128,0.008383110993438298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,2048,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,3072,32,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,51200,0.04648088746600681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,65536,0.05618933174345228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,65536,0.05207111106978523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,16384,0.017824000782436795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,65536,0.10899911324183147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,51200,0.04526933365397983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,12288,0.014354666074117025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,51200,0.08780355585945977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,16384,0.021361778179804485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,16384,0.018664888209766813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,10240,0.012451555165979596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,16384,0.03562044435077243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,12288,0.015127110812399121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,12288,0.02904711167017619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,10240,0.015728889240158927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,10240,0.013018666870064206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,8192,0.01033688916100396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,128,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,10240,0.026170666019121807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,7168,0.010048888623714447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,8192,0.011209777659840055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,8192,0.013765333427323235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,6144,0.008976889153321585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,7168,0.01033688916100396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,7168,0.012774222426944308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,7168,0.021643555826610986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,5120,0.007812444534566667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,6144,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,6144,0.01070577816830741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,51200,0.04115644428465102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,6144,0.019507555498017203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,5120,0.010066666536860997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,5120,0.017831110292010836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,4096,0.00867377801073922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,3584,0.007255111303594377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,4096,0.008279110822412703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,4096,0.015831111205948722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,3072,0.007920000288221572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,3584,0.007991111112965478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,3584,0.007637333538797166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,3584,0.014783999986118741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,2560,0.00703111125363244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,3072,0.007347555624114142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,3072,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,8192,0.022954665952258643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,2048,0.006173333360089197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,3072,0.014308444327778287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,2560,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,2560,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,1536,0.005038222091065513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,2560,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,2048,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,2048,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,4096,0.0075839997993575195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,1024,0.003944000022278892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,2048,0.012835555606418185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,5120,0.008987555901209513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,1536,0.006581333362393909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,768,0.003875555677546395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,1024,0.006226666685607698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,1536,0.011557333171367645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,12288,0.01719999975628323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,512,0.0031786666562159858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,1024,0.011056000159846412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,768,0.005440000030729506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,768,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,256,0.0024560000747442245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,768,0.011049778097205691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,512,0.005664000080691443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,512,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,512,0.010040889183680216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,256,0.005095110999213325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,256,0.010012444522645738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,64,0.0018604445374674266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,128,0.00509333320789867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,128,0.00555377784702513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,32,0.0017964444640609955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,128,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,65536,0.04994844396909078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,64,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,32,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,1536,0.005906666732496685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,65536,0.04988977644178602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,65536,0.05543377664354113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2560,1024,0.005548444473081165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,65536,0.10901955763498943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,16384,0.014518222875065274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,51200,0.044808000326156616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2560,128,0.002209777840309673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,256,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,12288,0.011719110939237805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,16384,0.018228444788191054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,16384,0.020629333125220407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,16384,0.03471822208828396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,10240,0.010320888625250923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,12288,0.014461333552996317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,12288,0.015232889188660515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,12288,0.029014223151736792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,10240,0.0124177775449223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,8192,0.009126221968068017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,10240,0.0151573336786694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,10240,0.025980444418059453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,7168,0.00851022203763326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,8192,0.011251555548773872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,51200,0.03943022092183431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,8192,0.01202933324707879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,6144,0.007996444072988298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,8192,0.023065777288542852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,7168,0.009999111294746399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,7168,0.012249777714411417
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,5120,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,7168,0.021254221598307293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,6144,0.009378666679064432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,6144,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,51200,0.08817244238323635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,4096,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,6144,0.019285332825448778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,5120,0.008639111287064022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,3584,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,5120,0.008813333180215623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,4096,0.008002666963471307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,5120,0.017542221479945712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,4096,0.00741244438621733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,4096,0.01588444411754608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,3584,0.007752888732486301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,3584,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,2560,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,3584,0.01477688882086012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,3072,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,3072,0.006970666348934174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,2048,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,3072,0.014470222923490735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,2560,0.007307555940416124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,2560,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,1536,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,2560,0.013423999978436364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,2048,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,2048,0.006274666637182236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,1024,0.004547555413511065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,2048,0.012433778080675336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,51200,0.03993955585691664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,1536,0.00656355544924736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,1536,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,1536,0.01237066669596566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,1024,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,1024,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,512,0.003173333282272021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,1024,0.011753777662913004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,768,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,256,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,768,0.01035733355416192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,512,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,3072,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,512,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,128,0.002187555655837059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,512,0.009899555808968013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,256,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,256,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,64,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,32,0.0018924443672100704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,128,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,128,0.004919111314747068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,65536,0.03532444437344869
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,128,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,64,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,2048,32,0.004922666483455234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,51200,0.028539554940329656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,65536,0.05466933382882012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,65536,0.04821866750717163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,16384,0.01240799989965227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,2048,768,0.003528888854715559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,65536,0.10857777463065253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,51200,0.04418666826354133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,51200,0.039248890346950956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,12288,0.009730666875839233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,16384,0.016626665989557903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,16384,0.017669333351982962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,768,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,10240,0.008985777695973715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,12288,0.013162666724787818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,12288,0.014708444476127625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,12288,0.028808888461854722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,8192,0.007832888927724626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,10240,0.01311288856797748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,10240,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,256,0.009706666900051964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,10240,0.025722665919197932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,7168,0.007998222278224098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,8192,0.010713777608341642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,8192,0.011021333436171213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,6144,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,7168,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,8192,0.022502221994929846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,7168,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,7168,0.02058844433890449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,6144,0.009710222482681274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,6144,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,6144,0.019070222973823547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,5120,0.008612444831265343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,51200,0.08717155456542969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,5120,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,5120,0.017271111408869427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,3584,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,16384,0.03471644388304816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,4096,0.007558222446176741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,4096,0.015808888607554965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,3072,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,3584,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,3584,0.007328888608349695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,3584,0.014807111687130399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,2560,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,3072,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,3072,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,3072,0.01384622189733717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,2048,0.005555555638339784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,2560,0.006979555719428592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,2560,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,5120,0.006168889088763132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,1536,0.004492444296677907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,2560,0.013791110780504016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,2048,0.00666844430896971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,2048,0.006142222219043308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,4096,0.006079111248254776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,2048,0.012442666623327466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,1536,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,1536,0.00562755556570159
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,768,0.0031822222388452957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,4096,0.007983110845088959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,1024,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,1024,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,512,0.0028515555378463534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,1024,0.011035555766688453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,768,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,768,0.0053075556125905775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,256,0.0023351110931899813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,768,0.010490666660997601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,512,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,512,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,128,0.0021404443929592767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,256,0.004952888935804367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,256,0.009693333672152625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,64,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,128,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,32,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,128,0.008294222255547842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,65536,0.025264889001846313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1536,1024,0.003549333247873518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,32,0.004896000027656555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,51200,0.020628444022602506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,65536,0.04645244280497233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,1536,0.012110222544935016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,65536,0.054177777634726636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,65536,0.10752889182832505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,16384,0.013282666603724161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,51200,0.03733955489264594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,51200,0.04401333464516533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,12288,0.010478222535716163
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,51200,0.08749866485595703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,16384,0.016581333345837064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,16384,0.015795555379655626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,16384,0.03461511267556085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,10240,0.009639999932712978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,512,0.00998577806684706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,12288,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,12288,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,8192,0.009116444322797988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,12288,0.028748446040683325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,10240,0.011761777930789523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,10240,0.011284444067213269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,7168,0.008569777839713627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,10240,0.02514044443766276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,8192,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,6144,0.007927999728255803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,8192,0.021820444199774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,7168,0.010042666561073726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1536,64,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,7168,0.009460444251696268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,5120,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,7168,0.019924443629052904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,6144,0.009399111072222391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,6144,0.008976000050703684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,4096,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,5120,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,6144,0.018224888377719455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,5120,0.01681866745154063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,4096,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,3584,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,4096,0.007317333585686154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,4096,0.01517777807182736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,3072,0.00554044461912579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,3584,0.007923555870850882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,256,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,3584,0.007116444408893585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,2560,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,3584,0.01478488908873664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,3072,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,3072,0.006903999795516332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,2048,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,2560,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,3072,0.014131555954615274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,2560,0.00655644428398874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,1536,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,2560,0.013056889176368713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,2048,0.007164444360468123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,2048,0.006239111224810283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,1024,0.003834666477309333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,1536,0.006569777925809224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,1536,0.00563733321097162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,5120,0.008746667040718926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,1536,0.011727111207114326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,768,0.0032115555885765287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,1024,0.0059164443777667145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,1024,0.005241777747869492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,512,0.0031537777847713898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,1024,0.011405333048767514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,768,0.005624889085690181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,768,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,256,0.0022026666750510535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,768,0.010642666783597736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,512,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,512,0.005088888936572605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,128,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,512,0.01000355515215132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,256,0.0053031109273433685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,256,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,64,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,8192,0.009708444277445475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,256,0.010070222119490305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,1024,32,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,128,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,65536,0.019903111788961623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,128,0.008233778178691864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,64,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,32,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,51200,0.016688000824716356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,2048,0.012330667012267642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,65536,0.05399822195370992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,65536,0.04576177729500664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,16384,0.012708444562223224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,65536,0.10775199863645767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,51200,0.037046220567491316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,12288,0.009905777871608734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,16384,0.01482488877243466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,51200,0.08640888664457534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,10240,0.00833333366447025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,16384,0.03460799985461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,12288,0.013105777402718862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,12288,0.01220533334546619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,8192,0.007351111206743453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,12288,0.02812888887193468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,10240,0.011622221933470832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,10240,0.011016888750924004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,7168,0.007274666594134436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,10240,0.025015999873479206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,8192,0.010437332921557956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,1024,128,0.004601777841647466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,8192,0.009761778016885122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,6144,0.006346666564544042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,8192,0.022022222479184467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,7168,0.009862222605281407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,5120,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,7168,0.020269332660569083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,6144,0.009457777771684859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,6144,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,4096,0.005971555494599872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,6144,0.018252443936136033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,51200,0.04336622357368469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,5120,0.008715555899673039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,5120,0.008084444536103142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,3584,0.005624889085690181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,5120,0.01714666684468587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,16384,0.015809777710172866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,3072,0.005409777578380372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,4096,0.015150222513410779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,3584,0.007931555310885111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,3584,0.007172444628344641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,2560,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,3584,0.015172445111804538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,3072,0.007361777954631382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,3072,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,3072,0.013799111048380533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,2560,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,2048,0.0042257776690853965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,2560,0.007268444531493717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,2560,0.013776889277829064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,7168,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,2048,0.006911110960774952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,2048,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,2048,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,1024,0.003479111111826367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,1536,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,1536,0.0057911111248864075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,1536,0.012091555529170565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,768,0.003162666741344664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,4096,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,1024,0.005245333330498801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,4096,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,512,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,1024,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,768,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,768,0.0052968888647026485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,256,0.0022026666750510535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,768,0.010381333529949188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,512,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,512,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,128,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,512,0.009697777529557547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,256,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,256,0.004931555440028508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,64,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,256,0.009655111365848118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,128,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,1536,0.00452533322903845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,128,0.004752000172932942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,65536,0.027634667025672063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,128,0.007687110867765214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,64,0.004930666751331753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,51200,0.021596444977654353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,768,32,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,65536,0.044891556104024254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,65536,0.050436443752712674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,16384,0.009018667042255402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,1024,0.005944889038801193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,65536,0.10781866974300808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,51200,0.038451555702421404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,12288,0.007466666400432587
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,51200,0.03660355673895942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,16384,0.015802666544914246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,51200,0.08614133463965522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,10240,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,16384,0.033596442805396184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,12288,0.012076444096035428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,12288,0.027753776974148218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,8192,0.005561777700980504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,10240,0.010716444088353051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,10240,0.01184088902340995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,7168,0.0053093334039052325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,10240,0.02437777817249298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,768,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,8192,0.009711999860074785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,6144,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,8192,0.021963554951879714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,7168,0.009127111070685917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,7168,0.019900444481107924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,6144,0.0090506664580769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,5120,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,6144,0.008641777767075432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,6144,0.018921777606010437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,4096,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,5120,0.00869511150651508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,5120,0.008084444536103142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,5120,0.01682577696111467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,3584,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,12288,0.013111111190583972
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,4096,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,4096,0.007287999822033777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,3072,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,4096,0.015843555331230164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,3584,0.007654222349325816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,3584,0.006912888752089606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,2560,0.004861333303981357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,3584,0.014863111906581454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,8192,0.010319111247857412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,3072,0.006974221931563483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,2048,0.004556444370084339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,3072,0.014429334137174817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,7168,0.0099982221921285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,2560,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,1536,0.003489777859714296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,2048,0.006412444429265128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,2048,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,2048,0.012737777498033313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,1536,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,1024,0.0031600000543726813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,16384,0.014339556296666464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,1536,0.011422221859296164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,768,0.002542222125662698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,1024,0.005420444326268301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,1024,0.011424000064531961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,768,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,3072,0.0069875551594628235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,768,0.005912888795137405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,512,0.002534222271707323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,768,0.011009777585665384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,2560,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,512,0.005622222191757626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,512,0.00488711107108328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,2560,0.013739555246300168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,128,0.0019662222928471034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,256,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,256,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,1536,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,256,0.010045333041085137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,64,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,512,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,128,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,65536,0.01625066664483812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,128,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,1024,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,64,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,512,32,0.0046942221621672315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,51200,0.013430222041077085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,65536,0.04241066508822971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,65536,0.04624799887339274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,16384,0.005987555616431766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,51200,0.03761511047681173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,65536,0.10666755835215251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,51200,0.034098668230904475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,12288,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,51200,0.08602933088938396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,16384,0.014117333624098035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,10240,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,512,0.010034666293197209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,12288,0.012059555285506778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,12288,0.02756266627046797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,10240,0.011692444483439127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,8192,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,128,0.005380444642570283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,10240,0.011094222466150919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,7168,0.0052435555391841466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,10240,0.02401599950260586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,8192,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,8192,0.009653333160612319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,6144,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,8192,0.02200711104604933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,7168,0.009684444301658207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,7168,0.019935111204783123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,5120,0.004570666700601578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,6144,0.009158222211731805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,6144,0.008622222476535374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,6144,0.018923555811246235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,4096,0.004531555705600315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,5120,0.008636444807052612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,16384,0.01581333412064446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,5120,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,3584,0.004611555486917496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,5120,0.016568889220555622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,4096,0.007941332956155142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,12288,0.01311911145846049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,3072,0.004561777744028303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,4096,0.007264888948864407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,4096,0.015247111519177755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,3584,0.006737777756320105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,2560,0.004571555389298333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,3584,0.014428445034556918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,3072,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,3072,0.014430221584108142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,2048,0.004536889079544279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,2560,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,2560,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,7168,0.009319999979601966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,2560,0.012773333324326409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,2048,0.0064533332155810455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,2048,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,2048,0.012088889049159156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,1024,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,1536,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,1536,0.011414222419261932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,16384,0.03388977713055081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,1024,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,768,0.002546666603949335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,1024,0.005994666781690385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,1024,0.010451555252075195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,3584,0.0075902218619982404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,512,0.0024764444679021835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,768,0.005730666634109285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,3072,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,768,0.004959110998445087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,256,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,768,0.01068800025516086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,512,0.005617777920431561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,512,0.004864000197913912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,128,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,512,0.010049777726332346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,256,0.00553955551650789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,64,0.0017866666118303935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,256,0.004598222259018156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,128,0.005429333282841577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,256,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,32,0.001793777777089013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,128,0.004755555755562252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,65536,0.011335999601417117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,128,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,64,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,1536,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,51200,0.009490666290124258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,65536,0.04599022203021579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,256,32,0.00481244424978892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,16384,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,65536,0.041618665059407554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,65536,0.10660444365607367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,51200,0.03733777668741014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,12288,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,51200,0.03394755721092224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,51200,0.08547377586364746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,16384,0.014138667119873894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,16384,0.015517334143320719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,10240,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,16384,0.03357155455483331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,12288,0.013078221844302284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,12288,0.012055111428101858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,8192,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,10240,0.011730666789743634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,12288,0.02781955401102702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,10240,0.01075644459989336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,7168,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,10240,0.024324445260895625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,8192,0.010336000058386061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,6144,0.004811555561092165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,8192,0.009641778137948778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,256,1536,0.003549333247873518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,7168,0.009702222214804756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,8192,0.02193244463867611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,7168,0.009014222357008193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,7168,0.020576889316240948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,6144,0.009331555830107795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,4096,0.004229333251714706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,5120,0.008662222160233391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,5120,0.00811111099190182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,5120,0.016528889536857605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,3584,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,4096,0.008106666306654612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,4096,0.0158595558669832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,3584,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,3072,0.004218666503826777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,3584,0.007673777639865875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,3584,0.015114666687117683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,3072,0.007540444533030192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,2560,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,3072,0.00655733338660664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,3072,0.014441778262456259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,5120,0.004531555705600315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,2048,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,2560,0.007255111303594377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,2560,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,6144,0.008361777497662438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,2560,0.012732444538010491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,6144,0.018412444326612685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,2048,0.006896889044178857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,1024,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,2048,0.005921777751710679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,1536,0.006523555351628198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,4096,0.007295110987292395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,2048,0.01274222218328052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,768,0.0024906665914588505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,1536,0.011363555159833698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,1024,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,1024,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,512,0.002464888824356927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,1024,0.011003555523024665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,768,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,768,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,256,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,768,0.010487111078368293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,512,0.005557333429654439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,128,0.001840888833006223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,512,0.005195555587609609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,512,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,256,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,64,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,256,0.005161777966552311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,32,0.0018088887963030075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,256,0.009013333254390294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,65536,0.006701333241330252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,128,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,51200,0.006552000012662675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,64,0.004950222041871813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,128,1536,0.0035004446076022256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,65536,0.041133334239323936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,12288,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,51200,0.033921778202056885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,10240,0.004729777574539185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,1536,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,16384,0.014146667387750415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,8192,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,7168,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,12288,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,10240,0.011008888483047485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,8192,0.009637333452701569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,7168,0.008974221845467886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,5120,0.004223111189073986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,6144,0.008275555239783393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,128,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,4096,0.003859555555714501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,128,0.004855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,5120,0.0076657773719893555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,3072,0.0038853333228164246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,16384,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,4096,0.007017778025733099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,2560,0.004397333496146732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,3584,0.006761777732107375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,2048,0.0041964443193541635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,1536,0.003540444291300244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,3072,0.006287999865081575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,1024,0.0028373334142896864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,2560,0.00610133343272739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,2048,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,768,0.002536000063021978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,512,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,1024,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,256,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,6144,0.004510222209824456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,128,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,768,0.004981333182917701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,512,0.004949333353175057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,64,0.0018115554832749898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,32,0.0018195555441909367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,256,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,65536,0.006551110910044775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,128,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,64,3584,0.004211555752489301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,64,0.0047493332790003884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,16384,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,12288,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,128,32,0.004889777965015835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,65536,0.0413955549399058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,10240,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,8192,0.003900444342030419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,51200,0.0333262218369378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,16384,0.01399466726515028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,12288,0.011759999725553723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,6144,0.00453599997692638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,10240,0.01037688884470198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,5120,0.0038942222793896994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,8192,0.009659555223253038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,1536,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,7168,0.0087013335691558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,3584,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,6144,0.008295999632941352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,3072,0.004184000194072723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,5120,0.007650666766696506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,2560,0.00423200014564726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,4096,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,2048,0.00424888895617591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,3584,0.006794666664467917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,1536,0.0031866667171319327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,3072,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,1024,0.0028355556229750314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,51200,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,768,0.0025128889828920364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,64,32,0.004578666554556953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,2048,0.006264888991912206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,512,0.0021848889688650766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,1536,0.005614222337802251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,256,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,128,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,1024,0.005342222336265776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,64,0.0016017777638302909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,768,0.005207111024194294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,7168,0.004253333227501975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,512,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,32,0.001791999985774358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,256,0.004580444345871608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,128,0.004566222015354368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,64,0.004724444614516364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,4,32,4096,0.003972444269392225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,32,0.004544888933499654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,16384,0.3335839907328288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,12288,0.2525342305501302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,16384,0.18012711736891004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,51200,0.4852382342020671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,4,32,2560,0.006039111150635614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,10240,0.21187733279334175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,12288,0.19848177168104383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,51200,1.0884862475925021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,12288,0.10327733225292629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,10240,0.13049156135982937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,8192,0.17344266838497588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,10240,0.10002399815453424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,7168,0.14191110928853354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,8192,0.12079555458492702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,6144,0.12337066067589654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,51200,0.4378631114959717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,7168,0.06304355462392171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,5120,0.1049742235077752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,6144,0.05388977792527941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,4096,0.08732266558541192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,5120,0.07412177986568876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,16384,0.13782400555080837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,5120,0.046851555506388344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,3584,0.07192177904976739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,4096,0.03939022289382087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,3072,0.06193333201938205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,3584,0.037800888220469155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,3584,0.05545777744717068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,2560,0.052116443713506065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,3072,0.03543555405404832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,3072,0.046562665038638644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,2048,0.04196622305446201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,7168,0.09649066792594062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,2560,0.032518221272362605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,2560,0.04348533352216085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,8192,0.0884560015466478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,2048,0.03663555449909634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,1024,0.022481777601771887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,1536,0.023245332969559565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,1536,0.02458133300145467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,768,0.017528888252046373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,1024,0.020615999897321064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,1024,0.01755288905567593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,512,0.012753778033786349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,768,0.01519555515713162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,768,0.01665866706106398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,256,0.0076657773719893555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,512,0.013258666627936892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,512,0.012812444733248817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,128,0.005072000126043956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,256,0.010379555324713389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,256,0.010049777726332346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,64,0.0041911109454101985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,128,0.00797155582242542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,1536,0.03296800123320685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,65536,32,0.0042071110672420925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,6144,0.08427733182907104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,64,0.00813244448767768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,32,0.008770667016506195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,65536,4096,0.06753244664933947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,65536,0.515042675866021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,16384,0.25421778361002606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,65536,1.0891573164198134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,51200,0.38754667176140684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,128,0.010599111517270407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,51200,0.8483893076578776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,65536,2048,0.028743998871909246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,12288,0.18841421604156494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,10240,0.15682133038838705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,12288,0.12309066454569499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,65536,0.4501226743062337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,51200,0.33570133315192324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,10240,0.10603911346859402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,16384,0.11275555027855767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,16384,0.15261156029171413
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,12288,0.09649333026674058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,8192,0.09252266751395331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,7168,0.11240177684360081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,8192,0.058324442969428174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,6144,0.0977368884616428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,5120,0.088919997215271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,7168,0.05159733361668057
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,7168,0.07872000005510119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,10240,0.06832355260848999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,4096,0.07111022207472059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,6144,0.04484533270200094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,5120,0.03865244322352939
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,3584,0.06262577904595269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,4096,0.03223289052645365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,5120,0.060439997249179415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,3072,0.05429955654674106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,4096,0.055086221959855824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,3584,0.045992887682384916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,3584,0.03170133299297757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,3072,0.029676444000667993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,2048,0.036660446061028376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,3072,0.038217776351504855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,2560,0.03631110986073812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,1536,0.028241776757770117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,2560,0.027311111489931744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,1024,0.019952888290087383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,2048,0.024311110377311707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,2048,0.03070666723781162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,8192,0.12861600187089708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,1536,0.019811555743217468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,768,0.01479466590616438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,1536,0.020098666350046795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,1024,0.01777688827779558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,768,0.013561777770519257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,256,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,768,0.0147386673423979
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,512,0.011124444504578909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,128,0.003181333343187968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,6144,0.06972266568077935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,256,0.008291555775536431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,64,0.002528000002106031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,128,0.009318222602208456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,32,0.0026462222966882917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,128,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,2560,0.04548888736300998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,64,0.007948444121413762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,32,0.007958222594526079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,65536,0.32026399506462944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,65536,0.17219111654493544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,51200,512,0.01035733355416192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,51200,0.248452451494005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,65536,0.16254399882422552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,51200,512,0.011126221881972419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,16384,0.08335822158389622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,51200,0.12973511219024658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,256,0.009532444179058075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,12288,0.06429066922929552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,51200,0.12498222457038032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,16384,0.06021866533491346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,10240,0.05449955662091573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,16384,0.042225778102874756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,8192,0.04247644543647766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,12288,0.033334222104814314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,7168,0.03626311156484816
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,10240,0.03591377867592706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,10240,0.043122665749655835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,8192,0.03649066554175483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,6144,0.031494223409228854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,8192,0.024139554964171514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,7168,0.02201066745652093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,5120,0.026877333720525105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,7168,0.031472888257768415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,6144,0.019730667273203533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,4096,0.022520888182852004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,51200,1024,0.015478221906556023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,3584,0.01939377850956387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,5120,0.024013333850436743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,4096,0.01555377741654714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,3072,0.01718488832314809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,4096,0.02255288925435808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,3584,0.020555555820465088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,3072,0.01514488955338796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,2560,0.01464355488618215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,3072,0.016453334026866488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,2048,0.012190221912331052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,2560,0.014706666270891825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,2560,0.016194666425387066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,1536,0.009742221898502773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,12288,0.05404177639219496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,2048,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,1024,0.0075031113293435835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,1536,0.010790222220950656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,1536,0.011058666639857821
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,768,0.005888000130653381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,1024,0.008979555633332994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,1024,0.009728000395827824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,512,0.0042204442951414315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,768,0.008168000313970778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,256,0.002530666689078013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,512,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,768,0.007999999655617608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,128,0.0022026666750510535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,512,0.006649777707126405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,3584,0.015507555670208402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,256,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,64,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,16384,32,0.0021199999998013177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,256,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,128,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,128,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,64,0.005943111247486538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,32,0.00574933323595259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,65536,0.24134310086568198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,2048,0.013418667018413544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,65536,0.14803200297885472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,16384,0.06428711281882392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,51200,0.18663644790649414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,12288,0.04956622256173027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,65536,0.13518310917748347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,51200,0.11123822795020209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,16384,6144,0.028328001499176025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,16384,0.037455999188952975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,10240,0.04197600152757433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,16384,5120,0.01741066740618812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,16384,0.05532977647251553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,51200,0.10473599698808457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,12288,0.03358577688535055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,8192,0.036835554573271014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,7168,0.0329368876086341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,12288,0.042744888199700244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,6144,0.028740445772806805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,8192,0.021653332644038733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,10240,0.03660444418589274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,7168,0.019928000039524503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,5120,0.024923556380801733
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,8192,0.03231733375125461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,4096,0.020900444851981267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,6144,0.017879999346203275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,3584,0.01870577699608273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,6144,0.024715556038750544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,5120,0.016197333733240765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,3072,0.016151999433835346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,4096,0.014242667290899487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,4096,0.02055022286044227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,3584,0.014004444082578024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,2560,0.01444355481200748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,3584,0.017429333594110277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,3072,0.013574221895800697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,2048,0.01239733315176434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,3072,0.01482399966981676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,2560,0.012540444731712341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,1536,0.008061333662933773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,2048,0.011390222443474663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,2048,0.012877333495352002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,1536,0.009759999811649323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,1024,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,1536,0.010033778018421596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,768,0.005199999858935674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,1024,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,10240,0.02575911084810893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,512,0.0031582222630580268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,1024,0.008230222596062554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,768,0.0075431110130416016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,7168,0.0277511113219791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,256,0.002469333302643564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,768,0.007612444460391998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,128,0.0021760000122918025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,512,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,256,0.006301333506902059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,5120,0.021903999977641638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,64,0.0018728888697094387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,12288,32,0.0019022222194406723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,256,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,128,0.006502222269773483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,128,0.005315555466545953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,64,0.005418666534953647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,65536,0.20137866338094076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,2560,0.01480888823668162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,51200,0.15846221976810032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,65536,0.13760622342427573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,65536,0.12554044193691677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,16384,0.055890666113959424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,12288,0.042698668109046094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,51200,0.10676178005006577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,51200,0.10034222073025173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,16384,0.047397332059012524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,10240,0.036575112077924944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,12288,0.0328453348742591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,8192,0.03048355711830987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,12288,0.03715822100639343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,12288,512,0.007273777491516537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,10240,0.024089778463045757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,7168,0.027666666441493567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,10240,0.03127911024623447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,6144,0.02442666722668542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,8192,0.026825777358478967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,7168,0.018929777873886954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,7168,0.023613333702087402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,5120,0.020650666620996263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,4096,0.017839110559887357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,6144,0.017444445027245414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,6144,0.02069244450993008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,5120,0.015375999940766228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,3584,0.01640533407529195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,4096,0.01402933398882548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,5120,0.01867022282547421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,3072,0.01404088901148902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,4096,0.016744888491100736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,3584,0.013225778109497495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,2560,0.012549333274364471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,3072,0.012854222622182635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,16384,0.03559199968973795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,2560,0.011682666838169098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,2560,0.012426666915416718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,1536,0.007524444825119442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,2048,0.010119111173682744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,2048,0.011345778074529437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,1024,0.005780444376998478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,1536,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,8192,0.020950223008791607
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,768,0.003891555385457145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,1024,0.007687999970383114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,1024,0.007663110891977946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,512,0.0031591111587153543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,768,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,12288,32,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,256,0.0023564443820052678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,768,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,512,0.0069155556460221606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,128,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,512,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,256,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,64,0.0019048889064126545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,256,0.005375111268626318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,32,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,10240,128,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,10240,2048,0.009637333452701569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,128,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,3072,0.013357333011097379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,64,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,65536,0.16757778326670328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,51200,0.12623466385735407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,32,0.005429333282841577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,16384,0.04495199852519565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,65536,0.09177688757578532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,51200,0.06822311215930514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,16384,0.0366231103738149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,16384,0.02254666719171736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,12288,0.03460266523891025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,10240,0.026586666703224182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,12288,0.018439999885029264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,12288,0.029620445436901514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,10240,0.023970666858885024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,10240,0.017190222938855488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,8192,0.015841777125994366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,3584,0.014530667000346713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,7168,0.020269332660569083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,8192,0.020612445142534044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,6144,0.018350222044520907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,7168,0.01458755632241567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,7168,0.018333334061834548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,5120,0.014783999986118741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,6144,0.013012444807423485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,65536,0.07243555784225464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,6144,0.01645155582163069
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,4096,0.012739555703269111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,5120,0.01166311071978675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,51200,0.05686133437686496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,3584,0.011229332950380115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,5120,0.014849777022997538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,4096,0.010604444477293225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,3072,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,4096,0.013253333667914072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,2560,0.009012444151772393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,3584,0.010024888647927178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,3072,0.009692444569534725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,2048,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,3072,0.01070844464831882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,2560,0.0086666668454806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,1536,0.006241777704821692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,2560,0.010392000277837118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,2048,0.008278222547637092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,1024,0.004573333180612988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,2048,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,1536,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,768,0.0032293332947625052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,1536,0.007294221884674496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,1024,0.006728888799746831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,512,0.0026417778184016547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,1024,0.006572444405820634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,768,0.0063271112740039825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,256,0.0022479999396536085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,768,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,512,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,128,0.0021315556433465746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,512,0.005626666463083691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,10240,1536,0.008749333520730337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,256,0.005889777921968036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,64,0.0018604445374674266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,32,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,8192,128,0.005668444352017508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,128,0.004867555366622077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,64,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,65536,0.14605777793460423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,8192,8192,0.023440000083711412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,32,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,51200,0.116230222913954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,65536,0.08473422129948933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,16384,0.041144887606302895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,65536,0.06569511360592313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,12288,0.031566222508748375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,51200,0.05323466658592224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,51200,0.06226933002471924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,10240,0.027036445008383855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,16384,0.02165066699186961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,16384,0.033428443802727595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,8192,0.023377777801619634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,12288,0.017708443933063082
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,12288,0.026851556367344324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,10240,0.01578044394652049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,7168,0.017782222893502977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,10240,0.023196443915367126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,6144,0.015643555257055495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,8192,0.014072888427310519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,5120,0.013637333280510373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,7168,0.012846222354306115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,8192,0.020231111182106864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,6144,0.012069332930776807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,4096,0.011715555356608497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,7168,0.017351999878883362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,3584,0.01035199976629681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,6144,0.016192888220151264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,5120,0.01441511180665758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,4096,0.009826666778988307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,3072,0.009233777721722921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,3584,0.012128888732857175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,4096,0.012879110872745514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,2560,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,3584,0.011857777833938599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,3072,0.00793866647614373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,2560,0.00796888851457172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,2560,0.010428444378905825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,1536,0.005646222167544895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,2048,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,2048,0.008634666601816813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,1536,0.006989333364698622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,1024,0.0038986665507157645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,1536,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,1024,0.00630666646692488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,768,0.00300799993177255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,1024,0.006513777706358168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,8192,256,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,512,0.002573333266708586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,5120,0.011008888483047485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,768,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,256,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,768,0.0059279998143513995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,512,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,3584,0.008940444224410588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,128,0.0021386666016446217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,512,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,64,0.0018213333355055915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,256,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,256,0.005219555563396878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,7168,128,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,32,0.0018862223045693503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,3072,0.010312000082598792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,128,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,65536,0.12143377463022868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,64,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,51200,0.09503644704818726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,7168,32,0.005197333378924264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,16384,0.03617777758174472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,65536,0.07298133108350965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,12288,0.027783112393485174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,51200,0.0511084430747562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,51200,0.05619733201132881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,16384,0.029725333054860432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,10240,0.023936889237827722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,12288,0.017249777913093567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,12288,0.025447110335032146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,8192,0.020445333586798776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,10240,0.020957332518365648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,7168,0.01818933255142636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,8192,0.018941332896550495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,7168,2048,0.006955555743641323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,6144,0.015847999188635085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,7168,0.012498666842778524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,6144,0.011722666521867117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,5120,0.013920888304710388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,6144,0.01516888870133294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,5120,0.01034400032626258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,16384,0.02068355513943566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,5120,0.013442666994200813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,4096,0.011166222393512726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,3584,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,4096,0.009047110875447592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,10240,0.015221332510312399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,3072,0.00851644410027398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,4096,0.013069333301650153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,3584,0.008660444782839881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,8192,0.013804444836245643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,2560,0.007312888900438945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,2048,0.006296000132958095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,3072,0.007613333563009898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,2560,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,3072,0.011008888483047485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,1536,0.005344888816277186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,2560,0.00960711141427358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,7168,0.016406221522225272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,1024,0.003560888684458203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,2048,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,1536,0.0069653333889113525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,1536,0.00664533343580034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,65536,0.0632257792684767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,1024,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,512,0.002495999965402815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,768,0.006237333433495627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,256,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,768,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,512,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,128,0.001957333336273829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,512,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,256,0.005600000007285013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,64,0.0018213333355055915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,256,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,128,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,32,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,128,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,3584,0.011319110790888468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,64,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,6144,32,0.005193777796294955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,51200,0.08143822352091472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,65536,0.06098310814963447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,2048,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,65536,0.06607200039757623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,16384,0.031558222240871854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,12288,0.0248364441924625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,6144,768,0.003154666680428717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,51200,0.049257778459125094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,51200,0.05072089036305746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,10240,0.021299555897712708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,16384,0.02013511127895779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,16384,0.029409776131312054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,8192,0.01830933325820499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,12288,0.01679733395576477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,10240,0.014807111687130399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,12288,0.023955555425749883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,7168,0.01644444465637207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,10240,0.019460444649060566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,8192,0.01315199997689989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,6144,0.014524444937705994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,7168,0.012289777398109436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,5120,0.012721777790122561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,6144,0.010999999940395355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,6144,0.014771555860837301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,4096,0.009593778186374241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,5120,0.009674666656388177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,5120,0.013081777426931592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,3584,0.008630221916569604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,4096,0.008444444172912175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,4096,0.012033777932325998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,3072,0.007845333053006066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,3584,0.008082666330867344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,3584,0.01092711091041565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,2560,0.00701155596309238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,6144,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,3072,0.007872000336647034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,2048,0.005884444548024072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,3072,0.008969777988062965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,2560,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,1536,0.004552000098758274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,2560,0.00865955568022198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,2048,0.006919999917348226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,1024,0.003160888950030009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,8192,0.01885066595342424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,1536,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,768,0.0029226665695508323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,7168,0.016028443972269695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,1536,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,1024,0.006254222244024277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,512,0.0025439999169773525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,1024,0.006215111249023014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,65536,0.10459911160998875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,256,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,512,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,768,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,128,0.0021457777669032416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,512,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,256,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,64,0.0018071112119489247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,128,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,5120,32,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,256,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,128,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,65536,0.08903288841247559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,64,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,32,0.004886222382386525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,16384,0.026272000537978277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,65536,0.05864977836608887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,65536,0.06085599793328179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,12288,0.019574221637513902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,51200,0.0473271111647288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,51200,0.04607022139761183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,16384,0.019536000159051683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,10240,0.016699555847379897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,5120,2048,0.007923555870850882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,16384,0.02642222245534261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,8192,0.014670222997665405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,12288,0.015983111328548856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,7168,0.013343111508422427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,12288,0.02165599995189243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,8192,0.012434666355450949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,10240,0.01885511146651374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,6144,0.01202933324707879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,5120,768,0.006266666783226862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,8192,0.017961778574519686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,7168,0.01165422217713462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,5120,0.010958222051461538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,7168,0.01479644411140018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,6144,0.010643555886215635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,5120,0.008982222113344405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,6144,0.013487999637921652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,3584,0.010109333528412713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,5120,0.013117333253224691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,4096,0.008280889027648503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,3072,0.009516444471147325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,4096,0.011076444553004371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,2560,0.00831733312871721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,51200,0.06740978029039171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,3072,0.007660444411966536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,2048,0.007127111156781514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,3072,0.008346666892369589
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,2560,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,1536,0.005299555758635203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,2560,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,2048,0.006940444310506185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,1024,0.0041582224269707995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,2048,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,1536,0.0063688887490166565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,768,0.0038648889296584656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,10240,0.014387556248241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,1024,0.006226666685607698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,512,0.0035075553589397003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,1024,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,768,0.005943999936183293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,768,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,4096,0.009666666388511658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,512,0.005759999983840519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,512,0.0052764444715446895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,256,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,256,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,3584,0.008033778104517195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,64,0.0018906665758954154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,128,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,3584,0.010370666782061258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,65536,0.07758844561047025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,64,0.004936000125275718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,32,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,51200,0.06124799781375461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,4096,1536,0.00638488887084855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,65536,0.057090666559007436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,16384,0.02332266668478648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,51200,0.044017778502570264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,51200,0.04660711023542616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,12288,0.017995556195576985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,16384,0.019215111931165058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,12288,0.01585244470172458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,12288,0.020938666330443487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,10240,0.01554844445652432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,4096,128,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,32,0.0018764444523387486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,8192,0.013376889129479727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,10240,0.014279999666743808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,7168,0.011861333416567909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,8192,0.012145777543385824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,10240,0.018039110634062026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,6144,0.011341333389282227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,8192,0.01628888812330034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,7168,0.011015111373530494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,5120,0.009687999884287516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,7168,0.014442667365074158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,6144,0.010031999813185798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,65536,0.057929780748155385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,4096,0.0086666668454806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,6144,0.01309333327743742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,5120,0.008977777428097194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,4096,0.00832177781396442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,4096,0.010517333116796283
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,3584,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,16384,0.025047999289300706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,3584,0.009362666971153682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,2560,0.007735999921957652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,3072,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,2048,0.006984000404675801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,3072,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,2560,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,1536,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,2048,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,2048,0.0069848886794514125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,1024,0.003920888735188378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,1536,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,1536,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,768,0.0034417777011791864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,1024,0.006275555739800136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,1024,0.005678222411208683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,512,0.003177777760558658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,3584,0.010072888599501716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,768,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,768,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,5120,0.011368888947698804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,512,0.005928888916969299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,128,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,4096,128,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,512,0.005233777893914117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,64,0.001826666709449556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,256,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,128,0.005610666755172942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,2560,0.007328000333574083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,128,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,65536,0.06769244538413154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,64,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,51200,0.053287999497519604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3584,32,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,16384,0.019917334119478863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,65536,0.05470755696296692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,51200,0.0424560010433197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,51200,0.04592000113593208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,12288,0.016178665889634024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,16384,0.01889333294497596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,16384,0.024380443824662104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,10240,0.013771555489963956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,12288,0.015840000576443143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,8192,0.012092444631788464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,12288,0.01958933307064904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,256,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,7168,0.011083555718262991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,10240,0.017437333861986797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,8192,0.011720000041855706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3584,3072,0.00887288898229599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,6144,0.010094222095277574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,8192,0.01521511044767168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,5120,0.009134222235944536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,7168,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,6144,0.009512888888518015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,6144,0.011325333681371478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,4096,0.007920888562997183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,5120,0.008830221990744272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,5120,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,3584,0.007351111206743453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,4096,0.008151111503442129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,4096,0.008651555412345463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,3072,0.008664888640244802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,3584,0.007686222592989604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,3584,0.008034666379292807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,2560,0.007826666865083907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,3072,0.007311999797821045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,10240,0.013831111292044321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,3072,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,2048,0.006952889263629913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,2560,0.007174222005738153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,2560,0.007419555551475949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,1536,0.006280000011126201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3584,256,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,2048,0.006978666616810693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,1024,0.004200889004601372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,7168,0.01050222251150343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,1536,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,768,0.003680888977315691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,1024,0.006250666661394968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,1536,0.006128888991143968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,512,0.0031724443866146933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,768,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,256,0.0027057776848475137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,1024,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,768,0.005583999885453119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,512,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,128,0.0022355556074115965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,65536,0.05697421895133125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,512,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,256,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,64,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,3072,32,0.0018462222069501877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,256,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,3072,128,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,65536,0.056799113750457764
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,128,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,51200,0.04542666673660278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,64,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,32,0.004921777794758479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,16384,0.01788444485929277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,65536,0.05576088693406847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,12288,0.01367644468943278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,65536,0.051757332351472646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,51200,0.04519288738568624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,16384,0.018802666001849704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,51200,0.041117333703570895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,10240,0.012042666474978128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,8192,0.01038577738735411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,12288,0.015234667393896313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,16384,0.021410667233996924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,7168,0.009703111317422655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,12288,0.017833777599864535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,8192,0.011029333704047732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,3072,2048,0.006636444479227066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,10240,0.015476443701320224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,6144,0.008988444175985124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,5120,0.008185777399275038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,8192,0.013973333769374423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,7168,0.012418666647540199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,6144,0.009681777821646797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,4096,0.007287110719415877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,5120,0.008755555583371056
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,6144,0.010335110955768162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,3584,0.0069324444565508105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,5120,0.009479111267460717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,3072,0.007975999679830339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,4096,0.008339555727110969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,2560,0.00665511108107037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,3584,0.00793688909875022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,3072,0.007647111184067196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,2048,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,3584,0.0075457774930530125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,3072,0.007384888827800751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,1536,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,2560,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,2048,0.006912888752089606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,1024,0.004193777839342753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,2048,0.006335999816656113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,1536,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,768,0.0038968887594011095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,1536,0.00590133335855272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,10240,0.013090666797426013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,1024,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,768,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,256,0.002521777732504739
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,768,0.005330666485759947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,7168,0.010394666757848527
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,128,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,512,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,256,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,256,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,128,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,32,0.0018817778262827133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,128,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,4096,0.008636444807052612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,65536,0.048003554344177246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,64,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,2560,0.007252444823582967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,51200,0.03958311014705234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2560,32,0.00489688871635331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,16384,0.01423377792040507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,65536,0.05512266688876682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,65536,0.04997155401441786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,12288,0.011609777808189392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,51200,0.04481688804096646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,16384,0.018223111828168232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,51200,0.039678222603268094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,10240,0.010207110808955299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,16384,0.020237333244747586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2560,512,0.0032088889016045462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,8192,0.00962933318482505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,12288,0.015156444576051501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,10240,0.01242844429281023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,7168,0.008838222258620793
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,10240,0.014426666829321118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,8192,0.010729778144094678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2560,512,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,6144,0.00795911086930169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,7168,0.010326222413116032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,8192,0.012795555922720166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,7168,0.010799999866220685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,6144,0.009687110781669617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,4096,0.009027555584907532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,5120,0.008707555631796518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,6144,0.009025777379671732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,3584,0.00851111114025116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,5120,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,4096,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,3072,0.005897777775923411
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,3584,0.007720888488822513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,4096,0.007641777396202087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,2560,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,3072,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,3584,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,2048,0.005929777605666055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,2560,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,1536,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,2048,0.006948444578382704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,2560,0.006600889066855113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,1024,0.0038355555799272326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,2048,0.006139555739031897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,12288,0.014479110638300577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,1536,0.006589333216349284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,768,0.0034968890249729156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,1536,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,1024,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,1024,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,768,0.0053137776752312975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,5120,0.007248000138335758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,256,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,512,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,512,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,128,0.002141333288616604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,256,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,64,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,128,0.005567111074924469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,32,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,256,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,128,0.004879111217127906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,65536,0.03559022148450216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,51200,0.028755555550257366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,64,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,3072,0.006980444822046492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,2048,32,0.00489155575633049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,65536,0.054808888170454234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,16384,0.012329777909649743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,65536,0.048067556487189404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,51200,0.044291555881500244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,16384,0.01688888834582435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,51200,0.038394666380352445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,10240,0.008647999829716152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,12288,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,16384,0.01773422294192844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,8192,0.0076657773719893555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,2048,512,0.002850666642189026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,12288,0.014783110883500842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,7168,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,10240,0.012079111403889127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,2048,768,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,6144,0.007283555964628856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,10240,0.0123831108212471
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,5120,0.006243555496136348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,7168,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,8192,0.010374222364690568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,6144,0.009597332941161262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,4096,0.006127999888526068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,7168,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,6144,0.00871377769443724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,5120,0.00868711123863856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,3584,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,5120,0.007994666695594788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,4096,0.008311111066076491
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,3072,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,3584,0.007958222594526079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,4096,0.007634667058785756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,2560,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,3584,0.007277333074145847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,3072,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,2560,0.007278222176763747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,3072,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,1536,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,2560,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,12288,0.009691555466916826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,2048,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,1024,0.0035706667436493766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,2048,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,1536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,768,0.0031439999325407874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,1024,0.005976000179847081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,512,0.0028577778074476454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,1024,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,8192,0.01072799993885888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,768,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,256,0.002414222185810407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,512,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,128,0.002143111079931259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,512,0.004960000101062986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,256,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,64,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,128,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,256,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,128,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,65536,0.025070221887694463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,64,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1536,2048,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,51200,0.02027466727627648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,32,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,65536,0.05420533484882779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,16384,0.013103110922707452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,65536,0.046559111939536206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,12288,0.010797333386209277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,51200,0.04383555716938443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,16384,0.015834665960735746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,51200,0.037350220812691584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,10240,0.009683555199040307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,16384,0.01616533266173469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,12288,0.01311377767059538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,8192,0.009283555878533257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1536,1536,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,12288,0.012799111505349478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1536,768,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,10240,0.012085333466529846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,8192,0.01037333326207267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,10240,0.01107377807299296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,7168,0.009665778113736046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,5120,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,7168,0.009346666435400644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,6144,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,4096,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,5120,0.008674666285514832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,5120,0.008016000191370646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,3584,0.005551110953092575
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,4096,0.007995555798212687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,3072,0.005535111245181825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,4096,0.007334222396214803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,3584,0.007445333732499017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,2560,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,3584,0.007151111132568783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,3072,0.007055111229419708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,2048,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,3072,0.006959111326270633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,2560,0.0069448889957533936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,1536,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,2560,0.0064311110311084324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,2048,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,7168,0.008323555191357931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,1024,0.0038817777401871155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,2048,0.005992888990375731
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,6144,0.007680888805124495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,768,0.0038497779104444715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,1536,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,512,0.003177777760558658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,1536,0.0058853332367208265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,8192,0.010035555395815108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,256,0.00216977774269051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,768,0.005940444353553984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,6144,0.009308444129096137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,1024,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,512,0.005442666510740916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,768,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,64,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,512,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,32,0.0018097776919603348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,128,0.005222222043408289
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,256,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,65536,0.019848000672128465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,128,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,51200,0.016365332735909354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,64,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,1024,32,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,65536,0.054008887873755566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,65536,0.04581422275967068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,16384,0.013445333474212222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,51200,0.042859554290771484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,51200,0.03701333204905192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,16384,0.015834665960735746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,16384,0.014800000521871777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,10240,0.00888533310757743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,12288,0.012823999755912356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,12288,0.0122079998254776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,1024,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,10240,0.011687999798191918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,10240,0.011047110789351992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,1024,128,0.0021626665774318907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,7168,0.007243555453088548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,8192,0.010391111175219217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,1024,256,0.005311111195219888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,6144,0.00628355559375551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,8192,0.009701333112186855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,7168,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,5120,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,6144,0.009297777381208207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,7168,0.009310222334331935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,4096,0.005872888697518243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,5120,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,6144,0.00867022242810991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,3584,0.005581333405441708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,5120,0.008357333640257517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,4096,0.00792622235086229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,3072,0.005196444276306364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,4096,0.0074551113777690465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,2560,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,3072,0.007452444069915348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,3584,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,12288,0.010044444766309526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,2048,0.004211555752489301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,3072,0.0069057775868309875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,1536,0.00434044458799892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,2560,0.006476444502671559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,2048,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,8192,0.007363555332024892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,1024,0.003907555507289039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,1536,0.00591555568906996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,768,0.003165333428316646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,1024,0.005882666756709416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,1024,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,512,0.0028320000403457214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,768,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,768,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,256,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,512,0.00545155546731419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,512,0.005231110999981563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,128,0.0021271111650599376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,256,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,3584,0.007642666498819987
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,256,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,64,0.0018213333355055915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,768,32,0.0018355554590622585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,65536,0.027501333090994094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,2560,0.007246221933099959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,64,0.004943111290534337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,2048,0.0069617778062820435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,32,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,65536,0.04492533206939697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,65536,0.04788444439570109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,1536,0.006256888724035687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,16384,0.009100444614887238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,51200,0.0385484430525038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,12288,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,51200,0.036525332265430026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,16384,0.015792000624868605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,10240,0.006976000136799282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,12288,0.01311822235584259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,16384,0.014539556370841132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,8192,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,12288,0.01183377785815133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,10240,0.011754666765530905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,7168,0.00518222235971027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,8192,0.010724444356229572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,10240,0.010737777584128909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,6144,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,8192,0.009975999593734741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,5120,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,7168,0.009295110901196798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,6144,0.008296888735559251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,4096,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,768,128,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,5120,0.008330666356616551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,5120,0.008284444610277811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,3584,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,768,128,0.004880889008442561
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,3072,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,51200,0.022436444958051045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,3584,0.007340444458855524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,4096,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,2560,0.004927999857399198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,3584,0.007298666569921706
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,3072,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,2048,0.004521777646409141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,2560,0.006796444455782573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,2560,0.00628977765639623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,1536,0.0035253332720862497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,2048,0.006736889067623351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,2048,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,1024,0.0029164445069101122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,1536,0.0063662222690052455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,1536,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,7168,0.009698666632175446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,768,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,1024,0.005955555372767978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,512,0.0025048889219760895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,6144,0.009391999906963771
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,768,0.005580444302823808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,256,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,512,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,512,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,128,0.0018862223045693503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,256,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,256,0.004659555438492033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,64,0.0018053334206342697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,128,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,512,32,0.001869333287080129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,128,0.004564444637960858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,65536,0.016513778103722464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,64,0.004549333204825719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,32,0.004907555464241239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,51200,0.013791999883121915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,65536,0.046352889802720815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,65536,0.04212533434232076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,16384,0.006115555349323485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,51200,0.03729866610633002
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,1024,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,51200,0.03432800041304694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,12288,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,16384,0.015792000624868605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,16384,0.014112888111008538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,10240,0.005562666803598404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,12288,0.01310222182008955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,8192,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,12288,0.012096000214417776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,512,4096,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,10240,0.011796444654464722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,7168,0.004918222212129169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,10240,0.010859555668301053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,6144,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,8192,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,7168,0.009703999592198266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,5120,0.004587555511130227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,7168,0.009076444639099969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,6144,0.009334222310119206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,4096,0.004528000122971005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,6144,0.00867911097076204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,5120,0.008648888932334052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,3584,0.004936888813972473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,5120,0.007993777592976889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,4096,0.00830666638082928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,3072,0.0047439999050564235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,4096,0.007640000018808577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,3584,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,2560,0.004575110971927643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,3584,0.00699733363257514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,3072,0.007305777735180325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,2048,0.004524444540341695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,3072,0.007035555938879649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,1536,0.0035111109415690103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,2560,0.005967111223273807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,2048,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,768,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,2048,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,1536,0.006380444599522485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,768,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,1536,0.005912000106440649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,1024,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,512,0.0024746666765875286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,1024,0.005240889059172736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,768,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,8192,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,256,0.002152888932161861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,512,3072,0.0069919998447100324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,768,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,512,0.00554666668176651
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,128,0.0017973333597183228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,64,0.0018088887963030075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,256,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,32,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,128,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,256,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,128,0.00462400002612008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,64,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,32,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,65536,0.04585422078768412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,65536,0.04145866632461548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,256,2560,0.007121777368916406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,16384,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,51200,0.03737511237462362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,51200,0.03390044305059645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,256,1024,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,16384,0.015559110376569959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,10240,0.00535111129283905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,16384,0.014470222923490735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,12288,0.013068444199032254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,12288,0.012075555821259817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,8192,0.005254222287072076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,10240,0.011732444167137146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,10240,0.010825778047243753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,7168,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,8192,0.010384000009960597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,8192,0.009699555734793345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,65536,0.011047999891969891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,6144,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,7168,0.009663999908500248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,5120,0.004520888957712385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,7168,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,6144,0.009320889082219865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,51200,0.009304000271691216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,4096,0.004541333350870344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,6144,0.008072888685597314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,3584,0.004588444613748127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,4096,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,12288,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,3584,0.007651555869314406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,3072,0.004568888909286923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,3584,0.0069253332912921906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,2560,0.004223111189073986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,3072,0.007012444237867991
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,3072,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,2048,0.004595555365085602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,2560,0.006768888897365994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,2560,0.006252444452709622
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,256,512,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,2048,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,2048,0.0061048890153567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,1024,0.0028808888875775864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,1536,0.0063893331421746155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,1536,0.0057706667317284485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,768,0.0025351111673646504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,1024,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,5120,0.00868088917599784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,1024,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,512,0.0024826667375034755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,768,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,256,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,768,0.004967999955018361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,4096,0.007606222397751278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,512,0.005650666852792104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,128,0.0018097776919603348
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,512,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,256,0.005249777601824866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,64,0.0017884444031450485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,32,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,256,0.004607999904288186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,65536,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,128,0.004952888935804367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,64,0.004552888787455029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,32,0.005127111242877113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,128,1536,0.003478222423129611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,65536,0.04109955496258206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,12288,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,51200,0.033858666817347206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,10240,0.004837333328194088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,8192,0.004868444469239977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,16384,0.014113777213626437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,7168,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,12288,0.011808888779746162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,6144,0.004608889006906086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,10240,0.010967999696731567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,8192,0.009364444348547194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,5120,0.004272888931963178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,4096,0.0042239998777707415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,128,5120,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,3584,0.0041857775714662345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,7168,0.009292444421185387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,3072,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,6144,0.008319999608728621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,2560,0.00424177779091729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,4096,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,2048,0.0042204442951414315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,3,128,128,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,1536,0.0035093331502543557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,3072,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,51200,0.006551110910044775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,1024,0.002855111120475663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,2560,0.005969777703285217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,768,0.002528000002106031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,512,0.002461333241727617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,2048,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,16384,0.005542222410440445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,256,0.0021751111166344746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,1536,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,128,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,1024,0.005105777747101254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,64,0.0018453333112928604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,64,32,0.0018151110659042995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,768,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,512,0.005220444252093633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,256,0.004592888885074192
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,51200,0.006239999913507038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,128,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,64,0.004564444637960858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,16384,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,32,0.004899555610285865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,12288,0.004795555439260271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,65536,0.04184266593721178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,10240,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,51200,0.033322665426466204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,8192,0.003907555507289039
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,16384,0.014450665977266101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,5120,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,7168,0.0042364444169733255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,12288,0.012080888781282636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,6144,0.004291555533806483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,10240,0.011005333728260465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,5120,0.004187555362780889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,64,3584,0.0069351109365622205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,8192,0.00962044464217292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,3584,0.004200889004601372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,7168,0.00886044485701455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,6144,0.008361777497662438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,3072,0.0038328886859946777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,5120,0.007967111137178209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,2560,0.0041697778635554844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,2048,0.0038951109680864546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,3584,0.007110222346252865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,1536,0.0031946665710873077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,3072,0.006966222491529252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,65536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,1024,0.002854222224818336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,2560,0.006506666541099548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,768,0.0026071110947264563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,2048,0.005869333528810077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,512,0.002230222233467632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,256,0.0021555556191338436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,1536,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,768,0.0052684446175893145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,128,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,512,0.004604444321658877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,64,0.0017928888814316855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,32,0.0015999999725156361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,256,0.004945777770545748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,128,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,64,0.00452888881166776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,3,32,4096,0.003923555629120933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,32,0.00452622233165635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,4096,0.007239999870459239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,16384,0.33307822545369464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,16384,0.22352356380886504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,12288,0.2508293257819282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,16384,0.16278666920132107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,51200,0.538831975724962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,3,32,1024,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,51200,1.1187707053290474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,12288,0.10235466559727986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,10240,0.12809688515133327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,8192,0.17175732718573677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,7168,0.14065333207448324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,8192,0.07012889120313856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,51200,0.43846578068203396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,7168,0.09694844484329224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,6144,0.12242666880289714
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,7168,0.06638489166895549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,8192,0.11950843864017063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,5120,0.10412533415688409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,6144,0.06051199966006809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,4096,0.08604177501466538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,5120,0.07376977470186022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,5120,0.05439822210205925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,12288,0.15240800380706787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,3584,0.07121955686145358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,3072,0.06157333321041531
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,4096,0.06949688990910848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,4096,0.049010667535993785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,3584,0.056377775139278836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,3584,0.04439999990993076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,2560,0.05187466740608215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,2048,0.04261777798334757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,3072,0.040928890307744346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,3072,0.048362665706210665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,2560,0.03380889031622145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,1536,0.03164444367090861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,2560,0.03951466745800442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,1024,0.022673777408070032
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,2048,0.029316445191701252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,1536,0.02368444369898902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,768,0.017431111799346078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,1536,0.027796443965699937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,1024,0.018031999468803406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,512,0.012884444660610624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,1024,0.020607999629444547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,768,0.014923555983437432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,256,0.007531555162535773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,6144,0.10519644286897446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,512,0.012772444221708508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,512,0.01310222182008955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,256,0.01037599974208408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,256,0.010065777434243096
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,64,0.0041928887367248535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,10240,0.08724888828065659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,128,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,32,0.0041848888827694785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,64,0.008716444174448649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,32,0.008082666330867344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,2048,0.03649955656793382
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,65536,768,0.016737777325842116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,10240,0.21001156171162924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,65536,0.5100880199008518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,128,0.010339555641015371
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,65536,1.0794906616210938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,16384,0.25194843610127765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,51200,0.4456968837314182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,51200,0.8484257592095269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,12288,0.1877546707789103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,16384,0.11160977681477864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,16384,0.17700710561540392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,65536,128,0.00489688871635331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,10240,0.15617511007520887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,51200,0.3301226562923855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,12288,0.12280000580681695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,10240,0.1022604438993666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,12288,0.08196266492207845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,7168,0.11221067110697429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,8192,0.0953342252307468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,6144,0.09710577461454604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,7168,0.07955111397637261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,7168,0.05304888884226481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,5120,0.08817155493630303
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,6144,0.05030666788419088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,6144,0.07259111271964179
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,4096,0.07110133435991076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,65536,0.43556621339586044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,5120,0.05894311269124349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,3584,0.06257333358128865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,5120,0.04540355669127571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,4096,0.03970488905906677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,4096,0.05509600043296814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,3072,0.05459733472930061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,8192,0.12814311186472574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,3584,0.046055999067094594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,10240,0.08169866932762994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,3584,0.03758488761054145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,3072,0.033209777540630765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,3072,0.041449778609805636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,2048,0.036200887627071805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,1536,0.028372443384594385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,2560,0.027913779020309448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,8192,0.05825422207514445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,1024,0.020086222224765353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,2560,0.03186755710177951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,768,0.01484444406297472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,1536,0.019811555743217468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,2048,0.030478222502602473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,1536,0.02370400064521366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,512,0.010456889039940305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,1024,0.017084444562594097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,256,0.00629333323902554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,768,0.0144177774588267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,512,0.011395555403497485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,128,0.0032053333189752367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,512,0.011375111010339526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,256,0.009288888838556077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,64,0.002587555597225825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,256,0.008261333737108443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,128,0.009296889106432596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,32,0.002528000002106031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,128,0.007311111523045435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,51200,2560,0.045462220907211304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,32,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,51200,64,0.007932444413503012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,65536,0.3169448905520969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,65536,0.17338400416904023
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,2048,0.02405244443151686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,51200,0.24703288078308105
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,16384,0.08295022116767035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,1024,0.015474667151769003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,51200,0.17716533607906768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,12288,0.06356355879041883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,65536,0.15782222482893202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,51200,768,0.013843555417325763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,10240,0.05396355523003472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,16384,0.06532444556554158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,51200,0.11671200063493516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,16384,0.042591998974482216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,12288,0.03313688768280877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,8192,0.04253688785764906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,7168,0.03809422254562378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,6144,0.03323644399642944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,10240,0.04367022381888496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,8192,0.02405244443151686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,7168,0.02295733326011234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,5120,0.02835644284884135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,8192,0.03731733229425218
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,7168,0.03352355625894334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,4096,0.023876445160971746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,6144,0.021868444151348535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,3584,0.020563556088341605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,5120,0.02093155516518487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,6144,0.02888444397184584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,3072,0.01716711123784383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,4096,0.018989332848125033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,4096,0.021935110290845234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,3584,0.018552889426549275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,2560,0.014769777655601501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,3584,0.01920977731545766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,3072,0.016639999217457242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,2048,0.012328888807031842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,3072,0.017093333933088515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,2560,0.01477599971824222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,1536,0.009832888841629028
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,2560,0.014137778017255994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,2048,0.013395555317401886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,1024,0.007531555162535773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,2048,0.013920888304710388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,1536,0.010782221953074137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,768,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,10240,0.028515554136700098
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,1536,0.0118568887313207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,512,0.004187555362780889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,1024,0.00907911111911138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,12288,0.050514668226242065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,256,0.0024986666523747975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,768,0.008338666624493068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,512,0.007304888632562425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,768,0.007931555310885111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,128,0.0021742222209771476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,256,0.006623111251327727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,512,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,64,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,16384,32,0.0021662221600612006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,256,0.005961777849329843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,128,0.006621333460013072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,5120,0.024938666158252295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,128,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,64,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,65536,0.23938221401638457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,51200,0.18530221780141196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,65536,0.14779110749562582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,16384,0.06404799885219999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,32,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,65536,0.13332800070444742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,12288,0.049633777803844877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,51200,0.15265956189897326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,51200,0.10192177693049113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,10240,0.04149688945876227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,16384,0.05567466550403171
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,12288,0.029374221960703533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,16384,1024,0.009300444689061906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,8192,0.03638844357596503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,12288,0.04285422298643324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,10240,0.02569777766863505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,7168,0.032645334800084434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,6144,0.02829244401719835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,8192,0.021877333521842957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,10240,0.037759111987219915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,7168,0.020350222786267597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,8192,0.032602667808532715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,7168,0.02794755498568217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,6144,0.01942933268017239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,5120,0.018332444959216647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,4096,0.0207315550910102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,6144,0.025622222158643935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,3584,0.018294221825069852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,5120,0.02131555477778117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,3072,0.016107555892732408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,4096,0.016755556066830952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,4096,0.02015911042690277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,3584,0.016216889023780823
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,2560,0.013843555417325763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,3584,0.01757688820362091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,3072,0.014731556177139282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,2048,0.012108444339699216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,3072,0.01608000033431583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,2560,0.013384000294738345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,1536,0.007987555530336168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,1024,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,2560,0.013447110851605734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,16384,0.04298133320278592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,2048,0.012740444805887012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,1536,0.00980355590581894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,768,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,512,0.003208000005947219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,1024,0.008394666843944127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,1536,0.010845333337783813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,256,0.002455111179086897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,1024,0.00830488900343577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,768,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,128,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,512,0.007299555672539606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,768,0.0074720001882976955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,5120,0.024664888779322307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,64,0.0018844445132546956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,512,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,12288,32,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,128,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,128,0.0052666668262746595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,64,0.005748444547255834
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,12288,32,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,51200,0.15782754951053196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,65536,0.1361280017428928
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,16384,0.05570755402247111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,65536,0.12276266680823432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,51200,0.0965413318739997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,12288,0.042259554068247475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,16384,0.03451733456717597
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,16384,0.04890133274926079
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,2048,0.011171555353535546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,10240,0.03531377845340305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,12288,0.027412444353103638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,12288,0.0362115568584866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,8192,0.03014577759636773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,10240,0.026760889424218073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,10240,0.031918221049838595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,256,0.006251555350091722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,7168,0.02642400066057841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,6144,0.02347733411524031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,8192,0.02656800051530202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,7168,0.01961688862906562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,5120,0.020959110723601446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,7168,0.023278221487998962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,65536,0.20193778143988717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,4096,0.017283555534150865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,6144,0.01833599971400367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,6144,0.0221404449807273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,5120,0.017242666747834947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,3584,0.016751110553741455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,4096,0.01566488875283135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,5120,0.018311111463440787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,3072,0.014192000031471252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,51200,0.1243893305460612
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,3584,0.014825777875052558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,4096,0.016511999898486666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,2560,0.012740444805887012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,3584,0.015053333507643806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,2048,0.009640889035330879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,3072,0.01346666696998808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,1536,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,3072,0.013735111388895245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,2560,0.011745777395036487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,2048,0.01038933296998342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,1024,0.005865777946180767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,2560,0.0123511114054256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,2048,0.011232888533009423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,1536,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,1536,0.009336888790130615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,1024,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,512,0.0031742221779293488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,256,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,768,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,1024,0.007785778078767989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,8192,0.020676443974177044
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,512,0.006979555719428592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,768,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,64,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,256,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,512,0.005947555518812603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,32,0.0019511110666725372
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,128,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,256,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,128,0.005229333208666907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,65536,0.16799378395080566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,64,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,10240,32,0.005326222214433882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,51200,0.12636978096432155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,16384,0.04515999886724684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,65536,0.09095999929640029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,65536,0.0702986650996738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,12288,0.034684442811542086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,16384,0.02459022237194909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,51200,0.05537244346406725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,51200,0.09338844484753078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,10240,0.027404444085227117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,12288,0.01851733360025618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,12288,0.027789334456125896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,10240,0.016177778442700703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,768,0.004108444270160464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,7168,0.021197333931922913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,10240,0.024653333756658766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,8192,0.014535999960369535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,6144,0.01923733287387424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,8192,0.02104533380932278
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,7168,0.013625777430004544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,10240,128,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,5120,0.015200000670221118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,7168,0.019122666782803006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,4096,0.012712000144852532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,6144,0.013021333350075616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,3584,0.012097777591811286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,5120,0.012508444488048553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,4096,0.011387555963463254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,5120,0.014749333262443542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,3072,0.0101724440852801
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,4096,0.013781333135233985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,3584,0.0107351111041175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,2560,0.008992888861232335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,3584,0.012379555238617791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,3072,0.00981777740849389
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,2048,0.008088000118732452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,3072,0.011415111521879831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,2560,0.00904444439543618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,1536,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,2560,0.010068444742096795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,2048,0.00795022232664956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,2048,0.009342222577995723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,16384,0.034959110948774554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,1024,0.004549333204825719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,1536,0.007324444750944774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,8192,0.023568888505299885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,1536,0.00796622203456031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,512,0.0028497777465316984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,1024,0.006615110983451207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,768,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,256,0.002190222135848469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,768,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,512,0.00600622221827507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,128,0.0021217777911159727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,512,0.00591111100382275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,256,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,64,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,256,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,6144,0.017441777719391715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,32,0.0018737777653667664
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,128,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,65536,0.1435662243101332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,64,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,32,0.0052586665583981415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,51200,0.11389333671993679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,65536,0.08205866813659668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,16384,0.040707555082109236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,65536,0.06535910897784762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,8192,768,0.003378666730390655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,51200,0.05340088738335503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,12288,0.03216444452603658
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,16384,0.021923555268181696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,10240,0.026762665973769292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,16384,0.032877332634396024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,8192,1024,0.006632888896597757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,12288,0.01774577796459198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,8192,0.023023110296991136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,12288,0.027106665902667578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,7168,0.01799199978510539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,8192,0.014111111561457316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,128,0.00591466658645206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,8192,0.020900444851981267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,7168,0.01796533332930671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,5120,0.014031110538376702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,6144,0.011985777980751462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,6144,0.016906667086813185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,4096,0.01201600001917945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,5120,0.010905777414639791
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,5120,0.01443555619981554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,3584,0.01051111105415556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,4096,0.00998488896422916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,4096,0.012813333008024426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,3072,0.009324444664849175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,3584,0.009304888546466827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,3584,0.012055111428101858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,2560,0.008307555483447181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,3072,0.008301333420806462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,3072,0.011101333631409539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,2048,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,10240,0.01551555593808492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,2560,0.007623111208279927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,1536,0.005759999983840519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,10240,0.0233608881632487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,2560,0.009648000200589498
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,6144,0.016294222739007738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,1024,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,7168,0.013030221892727746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,768,0.0031991110493739447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,1536,0.006722666737106111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,2048,0.008656000097592672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,1024,0.006594666590293248
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,1536,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,1024,0.006586666736337874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,768,0.006117333140638139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,256,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,768,0.005939555664857228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,512,0.005913777897755305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,128,0.002030222159292963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,512,0.005628444254398346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,256,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,64,0.0018764444523387486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,32,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,128,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,256,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,51200,0.08668977684444851
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,128,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,65536,0.12051733334859212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,64,0.005193777796294955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,51200,0.09747644265492757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,7168,32,0.005205333232879639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,16384,0.0359191099802653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,65536,0.07365777757432726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,65536,0.0627982219060262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,12288,0.027911110056771174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,51200,0.05118044548564487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,51200,0.07271733548906115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,10240,0.023880889018376667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,16384,0.02090488870938619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,12288,0.01717866626050737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,16384,0.030346665117475722
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,2048,0.007275555696752336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,12288,0.02422399984465705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,7168,0.01800355480776893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,10240,0.015202666322390238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,7168,512,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,8192,0.013456888496875763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,10240,0.022072000635994807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,6144,0.015857777661747403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,8192,0.019380443625979953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,7168,0.012715555727481842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,6144,0.011727111207114326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,7168,0.01718311177359687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,5120,0.010674667027261523
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,6144,0.015790222419632804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,3584,0.009526222116417354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,5120,0.013470222552617392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,4096,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,3072,0.008381333616044786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,4096,0.012596444123321109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,3584,0.008637333081828224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,2560,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,3584,0.011188444164064197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,3072,0.00793422261873881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,2048,0.006395555618736479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,3072,0.010372444159454769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,2560,0.0073315559162033936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,1536,0.005411555369695027
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,2560,0.008879111044936711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,2048,0.007239999870459239
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,8192,0.020284444093704224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,2048,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,1536,0.006696888970004187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,768,0.0031644445326593188
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,1536,0.006991110742092133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,1024,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,512,0.0025840000145965153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,5120,0.01402844488620758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,1024,0.006258666515350342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,768,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,256,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,768,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,4096,0.010813333094120026
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,512,0.005950222412745158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,128,0.0019804444164037704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,64,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,256,0.005623111294375525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,32,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,128,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,256,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,128,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,65536,0.10520888699425592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,64,0.005055111315515307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,32,0.0052133335007561584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,65536,0.06657777892218696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,65536,0.06074488825268216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,16384,0.030760000149408977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,51200,0.061736000908745654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,12288,0.025116443634033203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,6144,1024,0.003646222253640493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,16384,0.020290666156344943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,10240,0.020688888099458482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,16384,0.02879022227393256
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,12288,0.0166320006052653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,8192,0.018405333161354065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,12288,0.022554667459593877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,7168,0.016542222764756944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,10240,0.014953777194023132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,8192,0.013146667016877068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,10240,0.021271111236678228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,6144,0.014327110515700446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,6144,512,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,7168,0.012290666500727335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,5120,0.012602667013804117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,7168,0.015402666396564908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,6144,0.011041777829329172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,4096,0.010023111270533668
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,6144,0.015412444869677225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,5120,0.009973333113723332
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,3584,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,5120,0.01314133322901196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,51200,0.08270755741331312
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,3072,0.007625777688291337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,3584,0.007990222010347579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,4096,0.012420444852775998
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,2560,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,3584,0.011052444577217102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,3072,0.007640889121426477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,2048,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,51200,0.04918400115436978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,1536,0.004680888934267892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,2560,0.007342221836249034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,2560,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,1024,0.0033928888539473214
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,1536,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,2048,0.007669332954618666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,768,0.002895111011134254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,1536,0.006999111009968652
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,1024,0.006268444574541516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,512,0.002530666689078013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,1024,0.006301333506902059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,768,0.005952000204059813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,8192,0.0177831103404363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,768,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,256,0.0021928888228204516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,512,0.00591822216908137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,128,0.0019822222077184254
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,256,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,256,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,4096,0.008919999831252629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,128,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,32,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,64,0.004984000076850255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,2048,0.007135999699433644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,65536,0.08728088935216267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,65536,0.058832890457577176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,65536,0.058396445380316846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,51200,0.0676435563299391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,512,0.005331555588377847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,51200,0.06637777884801228
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,16384,0.025892444782786902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,51200,0.047247999244266085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,12288,0.01990133358372582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,128,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,16384,0.01959911154376136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,10240,0.016733333468437195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,12288,0.016009777784347534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,16384,0.026416889495319788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,8192,0.014085332552591959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,12288,0.02167911165290409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,10240,0.01442488862408532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,32,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,7168,0.013399110900031196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,10240,0.020114666885799833
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,6144,0.011763555308183035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,8192,0.017536888519922893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,7168,0.01145155562294854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,5120,0.010849778023031024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,7168,0.015450666348139444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,6144,0.010388444695207808
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,6144,0.014452444182501899
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,4096,0.009293333523803288
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,5120,0.011911110745535957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,3584,0.010261333651012844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,4096,0.008266666697131263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,4096,0.01168444421556261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,3072,0.01071999967098236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,5120,64,0.001834666563404931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,3584,0.00794222205877304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,2560,0.009690666364298927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,3584,0.009645333720578088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,2048,0.0069120000633928515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,3072,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,2560,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,1536,0.005938666562239329
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,5120,3072,0.009071110851234859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,2560,0.007352888584136963
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,1024,0.004231111043029361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,2048,0.007600000335110559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,8192,0.012706666356987424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,1536,0.006631111105283101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,768,0.003887999802827835
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,1536,0.006626666833957036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,1024,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,512,0.0035031110876136353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,1024,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,768,0.005948444621430502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,256,0.002198222196764416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,512,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,768,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,128,0.001869333287080129
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,256,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,512,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,64,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,4096,32,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,256,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,128,0.005606222069925732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,128,0.00490488898422983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,65536,0.07777777645323011
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,64,0.005246222019195557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,51200,0.062071111467149526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,32,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,16384,0.023593778411547344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,65536,0.056779556804233126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,65536,0.05779822005165947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,12288,0.01790488925245073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,51200,0.04660177893108792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,4096,3072,0.008004444340864817
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,51200,0.060121774673461914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,16384,0.019152889649073284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,10240,0.015459555718633862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,8192,0.01331999980741077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,16384,0.025684444440735713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,12288,0.015823110938072205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,10240,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,7168,0.012494222157531314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,12288,0.021168889270888433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,6144,0.010911111202504901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,7168,0.01094044413831499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,7168,0.014123555686738757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,6144,0.00964355551534229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,6144,0.013778666655222574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,5120,0.010026666853162978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,5120,0.008957333034939235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,4096,0.008837333156002892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,5120,0.011815999945004782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,4096,0.008282666405042013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,3584,0.010101333260536194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,4096,0.010312889185216691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,2048,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,3072,0.009487999810112847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,3584,0.00795111142926746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,3584,0.009337777892748514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,8192,0.012323555847009024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,3072,0.007636444436179266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,8192,0.015561777684423657
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,2048,0.006233777850866318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,2560,0.007310222420427535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,5120,0.008988444175985124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,2560,0.007348444726732042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,1536,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,2048,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,2048,0.006924444602595435
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,1024,0.003941333128346337
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,1536,0.006602666858169768
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,768,0.003744889050722122
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,1536,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,1024,0.006284444282452266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,512,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,1024,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,768,0.005972444597217772
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,256,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,768,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,128,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,512,0.005619555711746216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,256,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,64,0.0018586667461527719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,128,0.005564444594913059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,256,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,32,0.0018079999006456798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,128,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3584,2560,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,65536,0.06581244203779432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,64,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,51200,0.05215466684765286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,32,0.0048942222363419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,65536,0.05678933196597629
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,16384,0.01960444450378418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,65536,0.05408711234728495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,12288,0.01614755557643043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,51200,0.05244177911016676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,16384,0.018907555275493197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,10240,0.013738666971524557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,16384,0.022446221775478784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,12288,0.015793777174419828
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,12288,0.020243555307388306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,8192,0.012123555772834353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,10240,0.013896889156765409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,7168,0.011071110765139261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,10240,0.017256889078352187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,8192,0.011718221836619906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,6144,0.010323555933104621
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,512,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,7168,0.010063110954231685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,5120,0.009176889227496253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,7168,0.012816888590653738
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,6144,0.009331555830107795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,3072,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,4096,0.007986666427718269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,5120,0.009001777403884465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,3584,0.0075937774446275495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,5120,0.010020444790522257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,4096,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,3072,0.009212444225947062
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,4096,0.008995555341243744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,3584,0.007812444534566667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,2560,0.007266666326257918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,3584,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,51200,0.045828445090187915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,3072,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,2048,0.006172444257471297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,3072,0.007324444750944774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,1536,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,2560,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,2048,0.0069306666652361555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,2560,0.006682666639486949
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,1024,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,1536,0.006605333338181178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,768,0.0038888889054457345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,1536,0.006261333409282897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,1024,0.006280888699822956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,512,0.003528888854715559
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,1024,0.005602666487296422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,768,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,256,0.0025066667132907445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,768,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,512,0.005643555687533484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3584,10240,0.01884888940387302
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,6144,0.012060444388124677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,512,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,256,0.005598222215970357
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,256,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,128,0.005621333503060871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,32,0.0018871112002266778
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,128,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,64,0.004937777916590373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,32,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,51200,0.045915553967158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,65536,0.05176266696718004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,65536,0.056040002240075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,16384,0.01788444485929277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,2048,0.006573333508438534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,51200,0.04512088828616672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,51200,0.05111911230617099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,12288,0.013810666898886362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,16384,0.01884622209601932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,3072,8192,0.016019556257459853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,10240,0.011983999775515662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,16384,0.021347555849287245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,8192,0.0107342220014996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,12288,0.015221332510312399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,10240,0.012473777764373355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,12288,0.01738755570517646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,7168,0.009648888475365108
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,10240,0.016354666815863717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,6144,0.009323555562231276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,8192,0.013731555806265937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,7168,0.010120000276300643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,5120,0.00794044468137953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,65536,0.05673955546485054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,6144,0.009728000395827824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,6144,0.012040889097584618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,5120,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,5120,0.009124444590674506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,3584,0.006903111106819577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,4096,0.007988444632954067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,4096,0.008685333033402761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,3584,0.007662222617202335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,3584,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,2560,0.007179555793603261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,3072,0.007627555893527136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,3072,0.007104888558387756
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,2048,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,2560,0.007099555598364935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,3072,128,0.0022239999638663395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,2560,0.006592000110281839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,1536,0.005352889084153705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,8192,0.01071733319097095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,2048,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,1024,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,2048,0.006309333360857434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,768,0.0038488888078265717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,1024,0.0062773335311147906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,1536,0.0058808889653947614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,512,0.003142222141226133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,1024,0.005583111196756363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,4096,0.007363555332024892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,768,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,256,0.002518222149875429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,768,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,128,0.0021831111775504216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,512,0.005612444546487596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,256,0.005592000153329637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,512,0.005212444398138259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,64,0.0018320000833935207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,3072,0.007904888855086433
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2560,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,256,0.005211555709441503
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,65536,0.04797955685191684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,128,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,51200,0.03828977876239353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,64,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,32,0.00479911102188958
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,65536,0.055362668302324086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,65536,0.0496711101796892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,16384,0.014974221587181091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,12288,0.012098666694429187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,51200,0.04468355576197306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,51200,0.04920533299446106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,16384,0.018218666315078735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,10240,0.010019555687904358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,12288,0.013737777868906656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,16384,0.018541332748201158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,8192,0.009337777892748514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,1536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,12288,0.016911110944218107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,10240,0.012548444171746572
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,7168,0.00867555538813273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2560,7168,0.01204355557759603
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,6144,0.008000888758235509
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,10240,0.013916444447305469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,8192,0.010658666491508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,5120,0.007607111500369177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,7168,0.010055111514197456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,8192,0.012058667010731168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,4096,0.008619555168681675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,7168,0.010671111444632212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,3584,0.008678221868144141
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,5120,0.00865955568022198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,4096,0.008289777570300633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,5120,0.008619555168681675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,128,0.005576000031497743
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,3072,0.00628266649113761
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,4096,0.007663999994595845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,2560,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,3584,0.007939555578761632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,3584,0.007342221836249034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,3072,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,2048,0.007293333609898885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,2560,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,3072,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,1536,0.005196444276306364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,2048,0.0070017774899800616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,2560,0.006792888873153263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,1024,0.003936888857020272
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,2048,0.006113777971929974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,768,0.003891555385457145
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,1024,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,1024,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,512,0.0029004443850782183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,768,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,768,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,256,0.002479111154874166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,512,0.00563822231358952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,6144,0.00945955514907837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,128,0.0021546667234765156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,512,0.00527555536892679
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,6144,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,256,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,64,0.0018657777044508194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,2048,32,0.0018479999982648427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,256,0.004906666775544484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,128,0.005582222094138463
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,65536,0.03571466604868571
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,128,0.00490399988161193
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,51200,0.02880711025661892
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,64,0.004916444420814514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,32,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,65536,0.0547342234187656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,16384,0.012171555724408893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,65536,0.04831733306248983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,12288,0.010036444498433007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,51200,0.04425599839952257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,16384,0.017403556240929496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,51200,0.04319733381271362
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,10240,0.008719111482302347
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,16384,0.01759822169939677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,12288,0.013117333253224691
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,8192,0.007619555625650618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,12288,0.014811555544535318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,10240,0.012104888757069906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,10240,0.012582222620646158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,8192,0.010679999987284342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,8192,0.009885333478450775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,7168,0.010043555663691627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,7168,0.009676444033781687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,5120,0.006253333141406377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,6144,0.008775999976529015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,4096,0.006191111273235745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,5120,0.008660444782839881
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,5120,0.008302222523424361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,3584,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,4096,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,4096,0.00758844448460473
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,3072,0.005547555370463266
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,1536,0.006601777755551868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,3584,0.007301333049933116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,2560,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,2048,1536,0.005919111271699269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,3072,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,2048,0.005405333307054307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,7168,0.007970666719807519
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,3072,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,2560,0.007271111011505127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,1536,0.004977777600288391
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,2560,0.006562666760550604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,6144,0.0070186663005087115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,2048,0.006958222223652734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,1536,0.006647999915811751
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,2048,0.006086222413513396
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,768,0.003544888976547453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,1536,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,512,0.0028773333049482773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,6144,0.009392889009581672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,1024,0.005550222264395819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,256,0.0021395554973019496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,768,0.005292444593376583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,512,0.005643555687533484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,128,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,512,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,256,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,256,0.004866666677925321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,64,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,32,0.0018044445249769424
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,128,0.005577777822812398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,65536,0.024889777104059856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,3584,0.007677333222495184
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,64,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,32,0.005104889058404499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,51200,0.020257777637905545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,65536,0.05414933297369215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,65536,0.04651022288534376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,16384,0.013407111167907715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1536,1024,0.003670222229427762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,51200,0.041675554381476514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,12288,0.011359111302428775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,1024,0.006254222244024277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,16384,0.01585422290696038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,10240,0.009311111436949836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,768,0.005892444401979446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,12288,0.013329777452680798
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,8192,0.009352000223265754
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,12288,0.012662221988042196
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,10240,0.011832888755533429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,7168,0.008671999805503422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,10240,0.011380444798204633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,8192,0.010375111467308467
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,6144,0.006908444480763541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,8192,0.010048000348938836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,7168,0.010249777800507015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,5120,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1536,128,0.004830222162935468
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,7168,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,4096,0.005888888819350137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,6144,0.009340444372759925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,5120,0.00868000007337994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,6144,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,3584,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,4096,0.008344000412358178
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,5120,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,3072,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,4096,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,3584,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,51200,0.04383377896414863
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,3584,0.007142222589916653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,3072,0.007659555309348636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,2048,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,3072,0.006963555183675554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,2560,0.007320889168315464
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,1536,0.00453333349691497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,16384,0.015895111693276297
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,2560,0.00656088896923595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,2048,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,1024,0.0038533334930737815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,768,0.0035013332962989807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,2048,0.006276444428496891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,1536,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,512,0.002871111035346985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,1024,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,1536,0.005576888720194499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,256,0.002263111165828175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,768,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,1024,0.0053288886944452924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,768,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,512,0.005200888961553574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,64,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,256,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,32,0.0018782222436534034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,128,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,128,0.004892444445027245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,65536,0.01978488928741879
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,64,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,51200,0.016171556380059984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,1024,32,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,2560,0.004884444591071871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,16384,0.013321778012646569
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,65536,0.04550666610399882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,51200,0.04285066657596164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,12288,0.009767110976907942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,51200,0.037243554989496865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,16384,0.015823110938072205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,10240,0.008563555777072906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,16384,0.014569777581426831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,12288,0.013087111214796701
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,10240,0.012048000262843238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,10240,0.011023999916182624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,7168,0.006973333656787872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,512,0.00555288874440723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,8192,0.010687111152542962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,8192,0.009713778065310584
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,6144,0.006583111153708563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,7168,0.009768000079525841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,6144,0.00943288869327969
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,7168,0.009314667019579146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,6144,0.008639111287064022
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,4096,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,5120,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,5120,0.008185777399275038
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,65536,0.05405866768625048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,3584,0.005608888963858287
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,4096,0.0076044441925154785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,3072,0.0052177777720822235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,4096,0.0074275558193524676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,3584,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,2560,0.005221333354711533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,3584,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,3072,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,2048,0.004237333519591225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,3072,0.006913777854707506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,2560,0.007282666862010956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,12288,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,1536,0.004548444516128964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,2560,0.006593777901596493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,2048,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,1024,0.00416711096962293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,1536,0.006595555692911148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,1536,0.005698666804366642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,768,0.003091555502679613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,5120,0.005377777748637729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,1024,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,512,0.0028702221396896574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,1024,0.005578666511509154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,768,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,256,0.0021911110315057966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,512,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,512,0.005234666582610872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,128,0.0021484444538752237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,256,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,256,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,64,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,128,0.005609777652555042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,32,0.0018399999373488957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,128,0.00490133340160052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,768,8192,0.007269333634111617
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,65536,0.027059555053710938
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,64,0.004865777575307422
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,51200,0.02234755622016059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,32,0.004914666629499859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,1024,128,0.002165333264403873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,2048,0.005966222120655908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,16384,0.009072889056470659
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,65536,0.04513688882191976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,12288,0.007852444218264686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,51200,0.03836444351408217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,16384,0.01554666625128852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,51200,0.0365973346763187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,10240,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,16384,0.01421333352724711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,12288,0.012308444413873883
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,10240,0.011728000309732227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,768,768,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,10240,0.011038222246699862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,8192,0.010695999695195092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,6144,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,8192,0.009647111097971598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,7168,0.009696000152164036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,5120,0.0052560000783867305
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,7168,0.008928000099129146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,6144,0.009033777647548253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,4096,0.005630222045713001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,6144,0.008777778181764815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,5120,0.008375999828179678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,3584,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,5120,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,4096,0.008241777618726095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,4096,0.00700444479783376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,3072,0.004934222333961063
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,65536,0.048319111267725624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,3584,0.0075902218619982404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,3584,0.007019555403126611
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,3072,0.0074311114019817775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,2048,0.004598222259018156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,3072,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,2560,0.0069715554515520734
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,2560,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,8192,0.005208888815508948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,2048,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,1024,0.0031511110977994073
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,7168,0.0052373334765434265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,1536,0.006577777779764599
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,768,0.0028284444577164124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,1536,0.005552000055710475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,1024,0.006048000107208888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,512,0.0024906665914588505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,1024,0.0052284445199701525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,768,0.005587555468082428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,256,0.0022026666750510535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,512,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,512,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,128,0.001818666648533609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,256,0.0052151112920708126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,256,0.004596444467703502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,64,0.001810666587617662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,2560,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,128,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,32,0.0018035556293196147
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,65536,0.015850666496488783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,128,0.004570666700601578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,64,0.004523555437723796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,12288,0.01315822203954061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,32,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,65536,0.045888000064426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,16384,0.005970666805903117
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,2048,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,51200,0.03741333219740126
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,12288,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,16384,0.015844444433848064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,51200,0.03460977805985345
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,16384,0.014144889182514615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,10240,0.005379555539952383
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,12288,0.013095999757448832
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,512,768,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,8192,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,12288,0.012039999994966718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,10240,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,7168,0.0052533331844541765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,10240,0.010818666881985135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,8192,0.01032266683048672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,6144,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,8192,0.009715555442704095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,7168,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,7168,0.009306666751702627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,6144,0.009082666701740688
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,6144,0.008340444829728868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,4096,0.004531555705600315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,512,1536,0.0035439998739295532
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,51200,0.014289778139856128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,3584,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,5120,0.008332444561852349
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,65536,0.042135109504063926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,3072,0.004564444637960858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,4096,0.007629333270920648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,3584,0.007865777446164025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,2560,0.0052044445441828836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,3072,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,3584,0.0069164443347189165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,2048,0.004218666503826777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,3072,0.006763555523422029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,2560,0.0069919998447100324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,1536,0.0035342222286595237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,2560,0.006291555447710886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,2048,0.006933333145247565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,1024,0.0028657778683635923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,2048,0.00572622236278322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,1536,0.006513777706358168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,768,0.0025155554629034465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,1536,0.005607999861240387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,1024,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,512,0.002518222149875429
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,1024,0.005566222386227713
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,5120,0.004554666578769684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,768,0.0058906666106647914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,256,0.0021502222451898786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,768,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,512,0.005625777774386936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,128,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,5120,0.008649778034951951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,256,0.0052888890107472735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,512,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,64,0.001842666624320878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,256,32,0.0017893332988023758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,65536,0.011067555182509951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,128,0.005541333307822545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,128,0.004920000003443824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,4096,0.006928000185224745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,51200,0.009384888741705153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,64,0.004567111117972268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,32,0.004543111142185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,16384,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,65536,0.0414319998688168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,51200,0.036983112494150795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,16384,0.01551644504070282
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,51200,0.03357155455483331
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,10240,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,12288,0.012950222525331708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,12288,0.012101333174440594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,8192,0.0052328887912962176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,10240,0.011878222227096558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,10240,0.010741333166758219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,7168,0.00489777781897121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,8192,0.01068800025516086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,6144,0.004605333424276776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,7168,0.009872000250551436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,8192,0.009687999884287516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,5120,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,7168,0.008671110702885522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,6144,0.009190222455395592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,256,256,0.00461777796347936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,4096,0.0039048890272776284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,6144,0.00869511150651508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,5120,0.00828888929552502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,3584,0.004527111020353106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,5120,0.008099555969238281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,65536,0.04617422156863742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,3072,0.0042035554846127825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,3584,0.007672000262472365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,4096,0.006983111302057902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,2560,0.004537777768241035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,12288,0.004910222358173794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,3072,0.00756177802880605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,2048,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,2560,0.007296000089910295
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,3072,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,1536,0.0035146665241983202
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,2048,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,1024,0.0028737777223189673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,2048,0.005994666781690385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,1536,0.005611555443869696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,1024,0.005956444475385878
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,1024,0.004942222187916438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,768,0.002513777878549364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,768,0.0059262220230367445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,512,0.002522666628162066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,768,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,512,0.0055440002017551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,4096,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,256,0.002144888871245914
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,512,0.005202666752868229
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,128,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,3584,0.007033777733643849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,256,0.004576888763242298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,16384,0.014096000128322177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,64,0.0018435555199782054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,128,32,0.0018373332503769135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,128,0.004908444566859139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,65536,0.006666666517655055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,1536,0.006587555425034628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,51200,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,64,0.004556444370084339
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,16384,0.005209777918126848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,65536,0.0414319998688168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,12288,0.00490044429898262
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,51200,0.033569776349597506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,10240,0.004580444345871608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,16384,0.014107555150985718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,8192,0.004527111020353106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,12288,0.012087999946541257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,7168,0.004224888980388641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,10240,0.010911111202504901
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,6144,0.004537777768241035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,8192,0.009323555562231276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,5120,0.003847111016511917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,256,0.005260444349712796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,4096,0.003563555578390757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,7168,0.009170666337013245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,3584,0.00424888895617591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,128,0.00527022240890397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,2560,0.0062871111763848206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,5120,0.007882666256692674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,4096,0.006950221955776215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,2560,0.0042053332759274375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,2048,0.00387999994887246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,128,32,0.004574222283230888
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,1536,0.0032213332338465583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,3072,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,1024,0.002814222127199173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,768,0.002508444504605399
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,2560,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,512,0.0021191111041439907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,2048,0.005704888867007361
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,1536,0.005589333259397083
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,256,0.001992888955606355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,1024,0.005247999810510212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,128,0.001791999985774358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,64,0.0017964444640609955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,768,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,32,0.001820444439848264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,512,0.004863111095296012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,65536,0.00656444455186526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,256,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,128,0.004563555535342959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,51200,0.006231110956933763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,64,0.004522666749027041
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,32,0.004591111093759537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,12288,0.004270222038030624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,65536,0.0411102208826277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,10240,0.0042133331298828125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,51200,0.03324888812171088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,16384,0.014104000396198697
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,12288,0.011730666789743634
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,7168,0.004238222208287981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,3584,0.0070088886552386815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,6144,0.0042568888101312844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,8192,0.009367999931176504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,5120,0.0038515557017591265
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,7168,0.008787555827034844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,4096,0.003726222034957674
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,6144,0.008606221940782335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,3584,0.004238222208287981
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,5120,0.00794311116139094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,3072,0.0038444445365005066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,4096,0.006945778098371293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,2560,0.004233777936961916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,16384,0.005264889034960005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,3584,0.00700800038046307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,2048,0.0042168887125121225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,3072,0.006295111030340195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,64,6144,0.008613333106040955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,1024,0.0028364445186323593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,2560,0.006278222219811545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,8192,0.003920888735188378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,768,0.002509333400262727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,512,0.002149333349532551
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,1536,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,256,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,10240,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,128,0.00179822225537565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,1024,0.005274666680230035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,768,0.004924444274769889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,64,0.001790222194459703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,512,0.004573333180612988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,32,0.0018213333355055915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,256,0.004512888689835866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,128,0.00452533322903845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,64,0.004559111264016894
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,32,0.004579555657174853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,32,1536,0.003535110917356279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,16384,0.3304639922247993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,2,64,3072,0.003909333298603694
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,2,32,2048,0.005953777581453323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,12288,0.2509733306037055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,51200,1.1163760291205513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,16384,0.16005333264668783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,10240,0.20876533455318877
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,12288,0.12506932682461208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,8192,0.17104621728261313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,10240,0.11041333278020223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,8192,0.09216622511545818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,7168,0.1409671174155341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,51200,0.43789156277974445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,6144,0.121888001759847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,7168,0.082841780450609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,16384,0.17882843812306723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,6144,0.07175555494096544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,12288,0.19136267238193086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,5120,0.10344800021913315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,4096,0.08509955803553264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,5120,0.06083289119932386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,10240,0.1294364399380154
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,51200,0.5292444229125977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,4096,0.0644035538037618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,4096,0.04978311061859131
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,3072,0.06226666768391927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,3584,0.04542488853136698
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,3584,0.05894133117463854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,2560,0.05297688974274529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,3072,0.05134133497873942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,8192,0.115612440639072
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,3072,0.03974489039844937
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,2048,0.042671110894944936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,2560,0.03347466720475091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,2560,0.0427724454138014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,1536,0.03298755486806234
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,2048,0.029100444581773546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,2048,0.03506933318244086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,1024,0.023229334089491103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,1536,0.02304800020323859
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,1536,0.027460444304678176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,768,0.018008000320858426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,1024,0.01771111124091678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,512,0.012681778106424542
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,3584,0.07184622022840711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,768,0.014827556080288358
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,256,0.007876444194051955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,7168,0.10516533586714004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,768,0.016162667009565566
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,512,0.012888888518015543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,128,0.005030222237110138
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,256,0.010303999814722273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,64,0.004190222256713443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,128,0.010446222292052375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,65536,32,0.0041884444653987885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,128,0.007861333588759104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,6144,0.08938222461276585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,64,0.00795733349190818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,5120,0.0767422252231174
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,32,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,1024,0.019640889432695176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,65536,0.5563742319742838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,512,0.012716444830099741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,65536,256,0.00963644435008367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,65536,1.0820240444607205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,51200,0.42701421843634707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,16384,0.1776684390174018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,12288,0.18737599584791395
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,10240,0.15531822045644125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,16384,0.12791911760965982
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,12288,0.09949511289596558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,65536,0.4466382132636176
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,51200,0.3410995536380344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,8192,0.12754578060574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,10240,0.08422399891747369
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,7168,0.11200443903605144
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,8192,0.0726755526330736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,8192,0.09668888648351033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,7168,0.08062489165200128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,10240,0.10271644592285156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,6144,0.09643466605080499
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,7168,0.06814933485454984
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,51200,0.8426453272501627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,5120,0.08785066339704727
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,4096,0.07145511441760592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,6144,0.057609776655832924
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,3584,0.06188088655471802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,12288,0.15253599484761557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,5120,0.06247555547290378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,4096,0.05451911025577121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,4096,0.040729777680502996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,3072,0.053865777121649853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,3584,0.03769866625467936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,3584,0.048764444059795804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,2560,0.04534488916397095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,3072,0.03287911083963182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,2048,0.036144889063305326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,3072,0.041738665766186185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,2560,0.028706666496064927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,2560,0.03597066799799601
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,1536,0.02811644474665324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,2048,0.02421422302722931
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,16384,0.24969421492682564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,1024,0.020030222005314298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,1536,0.019893333315849304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,768,0.01475644442770216
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,1024,0.015384889311260648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,1536,0.023336889015303716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,512,0.010393777655230628
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,6144,0.07220621903737386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,1024,0.017486222916179232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,256,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,5120,0.05069955521159702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,768,0.014097778333557976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,128,0.003173333282272021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,512,0.011388444238238864
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,256,0.009355555805895064
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,64,0.002524444419476721
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,256,0.007648888561460707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,128,0.009348444640636444
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,51200,32,0.002528000002106031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,128,0.006622222148709827
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,64,0.006959999601046245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,32,0.007318221860461765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,65536,0.3181111017862956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,51200,2048,0.02923555506600274
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,51200,0.2442586686876085
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,16384,0.08220977915657891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,65536,0.20234133137596977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,51200,0.16330399778154162
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,12288,0.06314488914277819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,51200,0.12222577465905084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,512,0.011019555230935415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,16384,0.051663110653559365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,10240,0.05494399865468343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,16384,0.06639822324117024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,12288,0.04046488801638285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,8192,0.041947556866539844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,10240,0.04718577861785889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,8192,0.031129777431488037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,7168,0.03617422117127313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,8192,0.03695644603835212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,12288,0.050535112619400024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,6144,0.03121244576242235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,5120,0.026731555660565693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,7168,0.02976089053683811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,7168,0.03252799974547492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,6144,0.02698844505680932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,6144,0.02881866693496704
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,5120,0.023052444060643513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,3584,0.0196088883611891
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,5120,0.02439022229777442
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,4096,0.019889778561062284
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,51200,768,0.01344622257683012
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,4096,0.02200800014866723
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,3584,0.01867822143766615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,2560,0.01459111107720269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,3072,0.01779022150569492
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,2560,0.015096889601813423
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,2560,0.015375110838148328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,2048,0.01312888910373052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,10240,0.036213331752353244
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,1536,0.009760888914267222
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,2048,0.013316444224781461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,2048,0.013467555244763693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,1536,0.011065777805116443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,4096,0.02346399923165639
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,1536,0.011715555356608497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,768,0.005942222144868638
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,1024,0.00904799997806549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,1024,0.008966222405433655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,65536,0.16629689269595677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,3072,0.017754667335086398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,768,0.008166222108734978
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,256,0.002537777854336633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,3072,0.016900445024172466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,512,0.006272888845867581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,256,0.006628444625271692
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,256,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,128,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,128,0.006603555546866522
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,64,0.002159111201763153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,32,0.0021688888470331826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,1024,0.007919111185603673
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,64,0.005277333160241445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,32,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,65536,0.23880977100796172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,512,0.004016889052258598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,768,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,51200,0.18398133913675943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,512,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,16384,3584,0.020607999629444547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,65536,0.13338222768571642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,16384,128,0.002159111201763153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,51200,0.14387199613783094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,12288,0.0494391123453776
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,51200,0.10887110895580715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,16384,0.05554222067197164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,10240,0.041929778125551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,12288,0.03570222192340427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,12288,0.043212443590164185
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,10240,0.03155288762516446
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,8192,0.035754667388068304
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,10240,0.03933599922392104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,7168,0.03219288918707106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,8192,0.026706667410002813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,6144,0.029138667715920344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,8192,0.03191466795073615
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,7168,0.028112000889248315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,5120,0.02474400069978502
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,6144,0.022231111923853557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,65536,0.17365511258443198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,4096,0.02107288936773936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,16384,0.06371733215120104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,5120,0.019778667224778067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,5120,0.022327999273935955
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,3584,0.018974221414989896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,16384,0.043769776821136475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,4096,0.019661333825853135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,4096,0.017535999417304993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,3072,0.016823111308945548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,2560,0.014888000157144336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,3584,0.01573955516020457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,3584,0.018467555443445843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,2048,0.012272888587580787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,3072,0.014752000570297241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,2560,0.013049778011110095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,3072,0.016082665986484952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,1536,0.008088000118732452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,2560,0.014463999205165438
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,1024,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,2048,0.013076444466908773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,2048,0.011417778001891242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,1536,0.009712888962692684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,7168,0.025768889321221247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,768,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,512,0.0032008888406885993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,1536,0.010654222634103564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,1024,0.008381333616044786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,256,0.0025173332542181015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,1024,0.007975111405054728
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,512,0.007254222200976477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,768,0.007332444190979004
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,6144,0.02490933405028449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,128,0.0021253333737452826
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,64,0.001960888918903139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,256,0.006617777877383762
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,512,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,12288,32,0.002185777864522404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,256,0.005600888695981767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,128,0.006281777802440856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,128,0.005232000102599462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,65536,0.20170222388373482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,32,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,12288,64,0.005239999956554837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,16384,0.0548986660109626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,51200,0.15762221813201904
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,65536,0.16549600495232478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,65536,0.1276408831278483
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,12288,0.04225066635343763
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,10240,0.03647199935383267
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,51200,0.12584088908301458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,51200,0.10460088650385539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,16384,0.03802311089303758
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,16384,0.04794755578041077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,8192,0.029655999607510034
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,12288,0.032476445039113365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,7168,0.02721422248416477
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,12288,0.038741333617104426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,10240,0.027867555618286133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,6144,0.024452444579866197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,8192,0.023975999818907842
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,7168,0.0229795558585061
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,5120,0.02113955550723606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,10240,0.03155555658870273
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,7168,0.023931556277804907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,4096,0.017888888716697693
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,6144,0.02065333392884996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,6144,0.02103288968404134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,3584,0.016131555040677387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,5120,0.018402665853500366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,3072,0.01497688889503479
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,4096,0.015860444969601102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,5120,0.018283555905024212
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,2560,0.013559111290507846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,3584,0.014692443940374585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,3584,0.015136889285511441
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,3072,0.013086222112178802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,2048,0.010138666464222802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,2560,0.011794666449228922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,3072,0.013808888693650564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,1536,0.008288000192907121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,2560,0.012072000238630505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,2048,0.01036888857682546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,1024,0.0058497778243488735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,2048,0.011297778122954898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,1536,0.008999110923873054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,1536,0.00942488925324546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,768,0.004055110944641961
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,1024,0.00797333319981893
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,512,0.0032008888406885993
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,1024,0.007297777467303806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,768,0.007236444287829929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,256,0.0024497778051429326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,768,0.0068986668354935115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,8192,0.026503110925356548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,768,0.007598222129874759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,512,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,128,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,64,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,512,0.006256000035338932
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,10240,32,0.002128888956374592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,256,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,128,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,4096,0.016599110431141324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,64,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,65536,0.16366310914357504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,10240,32,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,51200,0.12353421582116021
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,16384,0.04207999838723076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,65536,0.11505422327253555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,51200,0.08646221955617268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,51200,0.05788266658782959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,16384,0.03529955612288581
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,12288,0.020434666011068556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,10240,0.029550221231248643
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,10240,0.01867377758026123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,8192,0.024098666177855596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,128,0.006248000181383557
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,10240,0.02436888880199856
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,7168,0.02181511123975118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,8192,0.016480889585283067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,8192,0.02123022245036231
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,6144,0.018499554859267343
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,7168,0.018498667412334018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,65536,0.07023466957939996
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,5120,0.016484444340070088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,12288,0.034506665335761175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,6144,0.016748444901572335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,16384,0.024107555548350017
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,5120,0.01309333327743742
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,5120,0.0147724449634552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,3584,0.012272888587580787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,12288,0.02751111156410641
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,4096,0.01164177805185318
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,4096,0.013094222380055321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,3072,0.010018666585286459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,3584,0.010752889017264048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,256,0.006599111275540457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,2560,0.00902133352226681
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,3072,0.009689778089523315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,2048,0.007582222421964009
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,3072,0.0110871113008923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,2560,0.008851555486520132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,1536,0.006298666612969504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,2560,0.010589333044158088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,2048,0.008208889100286696
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,1024,0.004553777890072929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,2048,0.009347555538018545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,7168,0.01585244470172458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,1536,0.0074879998962084455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,768,0.0035031110876136353
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,1536,0.007656888829337225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,1024,0.006618666566080517
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,6144,0.014326221413082547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,768,0.006606222026877933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,1024,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,256,0.0023626666516065598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,768,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,128,0.0021768889079491296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,256,0.005616000129116907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,256,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,128,0.005740444279379315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,64,0.0018560000591807896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,128,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,3584,0.012672888735930124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,65536,0.1436328887939453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,64,0.004876444323195351
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,32,0.004888888862397936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,51200,0.11260355843438043
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,65536,0.09979289107852513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,65536,0.06653955247667101
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,51200,0.0862613320350647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,512,0.0028773333049482773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,51200,0.053103999959097974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,12288,0.031699554787741765
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,16384,0.021464889248212177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,8192,4096,0.013416888813177744
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,16384,0.03346044487423367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,10240,0.027143999934196472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,12288,0.017959111266665988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,8192,0.023265777362717524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,12288,0.026542221506436665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,10240,0.016415999995337594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,7168,0.018295110927687753
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,8192,0.014018666413095264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,8192,0.020421332783169217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,6144,0.015746666325463187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,7168,0.01295377810796102
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,7168,0.01815555493036906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,5120,0.013561777770519257
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,6144,0.01204533295498954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,6144,0.01610666679011451
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,4096,0.012056888805495368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,5120,0.011021333436171213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,16384,0.040735999743143715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,5120,0.013800889253616333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,3584,0.010616000327799054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,4096,0.009658666948477427
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,3584,0.008971555365456475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,4096,0.012478222449620565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,3584,0.011780444118711682
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,8192,512,0.005594666633341048
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,2560,0.008392888638708327
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,3072,0.008348444269763099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,2048,0.007135999699433644
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,3072,0.011413333316644033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,2560,0.00830577810605367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,10240,0.02251022226280636
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,1536,0.005879111174080107
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,2048,0.007261333366235097
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,1024,0.003960888832807541
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,1536,0.006949333681000604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,1536,0.007196444604131911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,768,0.003481777591837777
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,1024,0.006609777609507243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,1024,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,512,0.00282222218811512
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,768,0.006288888967699475
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,768,0.005886222339338726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,256,0.0021626665774318907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,512,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,512,0.0052844443255000645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,3072,0.009538667069541084
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,512,0.006215111249023014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,256,0.0059368887709246735
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,64,0.0018524444765514797
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,256,0.004935111022657818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,32,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,128,0.005604444278611078
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,128,0.004607111215591431
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,65536,0.11984354919857448
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,64,0.004932444542646408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,2560,0.009930666950013902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,32,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,51200,0.09551555580563015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,16384,0.035742223262786865
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,65536,0.08654399712880452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,65536,0.06292888853285047
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,51200,0.051035556528303355
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,12288,0.027854220734702215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,16384,0.020907556017239887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,10240,0.0240000006225374
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,16384,0.029821332958009508
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,12288,0.01718222267097897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,8192,0.020250666472646926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,12288,0.024558222956127588
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,10240,0.015164444843928019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,10240,0.020624889267815482
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,7168,128,0.0021155555215146807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,6144,0.016082665986484952
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,8192,0.018979556030697294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,7168,0.012699555191728802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,5120,0.014288889037238227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,7168,0.016398222910033334
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,6144,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,4096,0.010939555863539377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,6144,0.01556711064444648
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,5120,0.010604444477293225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,3584,0.00959822204377916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,4096,0.008987555901209513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,5120,0.013076444466908773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,7168,2048,0.008588444855478074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,3072,0.008527999950779809
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,4096,0.012306667036480375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,2560,0.0075777777367168
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,3584,0.008283555507659912
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,3584,0.011056889262464313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,2048,0.006695111178689533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,3072,0.010689777632554373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,2560,0.007303111255168915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,1536,0.005392000079154968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,2560,0.009390222529570261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,2048,0.007276444799370236
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,1024,0.003842666745185852
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,1536,0.006624889042642381
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,7168,0.017853332890404593
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,8192,0.01368711143732071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,1536,0.0069297779765393995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,1024,0.006267555471923616
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,1024,0.006242666807439592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,512,0.0027546667390399506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,768,0.005605333381228977
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,512,0.005920889063013925
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,256,0.0021840000732077495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,512,0.0052791109515561005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,256,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,3072,0.007931555310885111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,51200,0.07231911023457845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,128,0.005615111026499007
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,64,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,128,0.004564444637960858
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,2048,0.007653333246707916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,32,0.0018364443547195857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,768,0.0032071111102898917
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,65536,0.10553689135445489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,64,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,768,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,51200,0.08151911364661323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,65536,0.060765332645840116
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,16384,0.03070488903257582
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,51200,0.0492195553249783
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,51200,0.06692177719540067
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,12288,0.024738666084077623
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,16384,0.020081778367360432
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,6144,128,0.002096000024014049
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,16384,0.0292195545302497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,256,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,10240,0.020503110355801053
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,8192,0.01886755559179518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,12288,0.016543999314308167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,12288,0.023696889479955036
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,7168,0.016542222764756944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,10240,0.014863111906581454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,10240,0.02017688916789161
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,8192,0.013077333569526672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,6144,0.014430221584108142
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,7168,0.012424888710180918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,8192,0.017426666286256578
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,5120,0.012439111040698158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,7168,0.01660711069901784
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,6144,0.011211555865075854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,4096,0.010271111296282874
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,6144,32,0.0047102222839991255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,5120,0.009705777797434065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,3584,0.008566222257084316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,4096,0.008653333617581261
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,5120,0.012725333372751871
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,3072,0.007744889292452071
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,65536,0.08051021893819173
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,4096,0.01205599970287747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,2560,0.0069191112286514705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,3584,0.01106844428512785
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,3072,0.007683555285135905
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,2048,0.006085333310895496
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,3072,0.009670221971140968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,2560,0.0075804442167282104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,1536,0.004586666822433472
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,2560,0.008620444271299574
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,1024,0.0038417776425679526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,2048,0.007130666739410824
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,768,0.003185777821474605
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,2048,0.007615110940403408
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,1536,0.006721777634488211
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,512,0.0027057776848475137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,1024,0.0063075555695427795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,1536,0.006777777853939269
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,256,0.002181333386235767
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,1024,0.00590844452381134
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,512,0.005930666708283954
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,768,0.005616888817813661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,128,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,256,0.005569777968857024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,512,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,64,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,5120,32,0.0018417777286635505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,256,0.004913777940803104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,128,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,6144,0.014427555931939019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,128,0.004590222405062782
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,65536,0.08567555745442708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,51200,0.06712266471650866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,64,0.004873777843183941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,5120,32,0.004705777598751916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,16384,0.025663110944959853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,3584,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,65536,0.05860533316930135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,65536,0.07157688670688205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,12288,0.019123555885420907
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,10240,0.016752888758977253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,51200,0.04741511080000135
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,51200,0.06122222211625841
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,16384,0.01942399972014957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,8192,0.01497866710027059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,16384,0.02606044378545549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,7168,0.012795555922720166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,10240,0.014365333649847241
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,12288,0.021879111727078755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,8192,0.01261333293384976
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,6144,0.012409778104888068
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,10240,0.020947555700937908
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,7168,0.01165600038237042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,5120,0.010577778021494547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,768,0.005934222290913264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,6144,0.01014311114947001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,6144,0.013530666629473368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,4096,0.009292444421185387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,5120,0.009015111459626092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,3584,0.01072533345884747
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,5120,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,4096,0.008295111358165741
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,3072,0.010089777410030365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,4096,0.011371555427710215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,3584,0.007997333175606197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,2560,0.009119999905427298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,3584,0.010004444254769219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,3072,0.007631111476156447
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,3072,0.008376888930797577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,2048,0.008299555215570662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,2560,0.007231999602582719
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,1536,0.005666666560702854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,2560,0.007909333540333642
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,12288,0.016009777784347534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,1024,0.004648888690604104
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,1536,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,1536,0.00647555540005366
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,8192,0.01664977769056956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,768,0.0038488888078265717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,1024,0.00627377794848548
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,7168,0.014656888114081489
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,768,0.005971555494599872
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,768,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,256,0.0023333333018753263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,512,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,512,0.005227555417352253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,128,0.00184888889392217
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,256,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,256,0.004856889032655292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,64,0.001830222292078866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,128,0.004548444516128964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,32,0.0018213333355055915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,2048,0.0069475554757648045
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,64,0.004870222260554631
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,2048,0.0069208890199661255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,65536,0.0770053333706326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,32,0.004638222356637319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,51200,0.06096444527308146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,65536,0.057560000154707164
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,4096,1024,0.005917333480384614
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,65536,0.06830933358934191
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,16384,0.023230221536424425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,51200,0.059177776177724205
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,12288,0.01795733306143019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,51200,0.04661866691377428
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,10240,0.016182222300105624
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,16384,0.019096000327004325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,8192,0.013423999978436364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,12288,0.015803555647532146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,16384,0.025056888659795124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,10240,0.01423555612564087
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,12288,0.019896888070636325
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,8192,0.012235555383894177
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,10240,0.017528888252046373
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,6144,0.010997333460383944
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,8192,0.016167110866970487
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,7168,0.010766222245163388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,5120,0.00977955593003167
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,7168,0.014157333307796054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,6144,0.009855111440022787
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,4096,0.008667555948098501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,6144,0.012763555679056378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,3584,0.010464888479974536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,5120,0.010795555180973478
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,4096,0.00832711077398724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,3072,0.009520889156394536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,4096,0.009522666533788046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,3584,0.007976888782448238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,2560,0.007607999775144789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,4096,512,0.003216888962520493
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,3072,0.007657777931955125
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,2048,0.007015110717879401
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,3072,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,2560,0.007308444215191736
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,2560,0.007225777539942
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,1536,0.005941333456171884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,2048,0.007045333584149678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,1024,0.004385777645640903
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,2048,0.006523555351628198
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,7168,0.012703999876976013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,1536,0.006611555400821898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,768,0.003863111138343811
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,1536,0.005954666684071223
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,1024,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,128,0.005572444448868434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,1024,0.005374222166008419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,768,0.005919999960396025
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,256,0.0021911110315057966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,768,0.0052639999323421055
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,512,0.0059057776298787855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,128,0.0018337778747081757
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,5120,0.009015999734401703
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,256,0.005629333357016246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,64,0.0018613334331247543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,256,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,128,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,32,0.001847111102607515
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,128,0.004538666870858935
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,65536,0.06821244292789035
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,3584,0.008988444175985124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,64,0.0047422221137417685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,51200,0.0528053343296051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,32,0.004588444613748127
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,16384,0.02013422217633989
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,65536,0.056792888376447886
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,51200,0.045788443750805326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,51200,0.05639555719163683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,12288,0.016519111063745286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,16384,0.018935999936527677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,16384,0.022625777456495497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,12288,0.01553244392077128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3584,512,0.003149333306484752
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,8192,0.01202311118443807
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,10240,0.013824889229403602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,10240,0.01607733302646213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,7168,0.01111288865407308
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,8192,0.011380444798204633
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,8192,0.015022221538755627
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,6144,0.010372444159454769
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,7168,0.010262221925788457
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,7168,0.012872888810104795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,5120,0.009284444153308868
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,6144,0.009695111049546136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,6144,0.0120328888297081
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,4096,0.007925333248244392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,65536,0.06948177682028876
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,5120,0.008774222599135505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,5120,0.010304888917340172
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,3584,0.007300444775157505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,4096,0.008281778130266402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,4096,0.008897778060701158
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,3072,0.009248889154858058
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,10240,0.014144000079896716
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,2560,0.008223999705579545
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,3584,0.00795377790927887
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,3072,0.007845333053006066
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,12288,0.019368888603316415
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,2048,0.0069217777086628815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,3072,0.007626666790909237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,2560,0.007270221908887227
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,1536,0.0058959999846087555
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,2560,0.00665155549844106
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,2048,0.007089777953094906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3584,512,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,1024,0.0046675557063685525
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,768,0.0038977778620190094
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,1536,0.006599999964237213
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,2048,0.006584888945023219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,512,0.0033697777738173804
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,1024,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,1536,0.00592533333433999
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,256,0.0025111111915773815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,768,0.005910222315125995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,1024,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,128,0.002326222136616707
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,512,0.005893333504597346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,256,0.005571555760171678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,512,0.004872000051869286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,64,0.0018151110659042995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,3072,32,0.0018568889548381169
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,128,0.005559999909665849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,256,0.004819555415047539
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,65536,0.056999113824632436
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,128,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,51200,0.04606399933497111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,64,0.004568888909286923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,16384,0.01767200065983666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,32,0.004568888909286923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,3584,0.007965332931942409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,65536,0.056127998563978404
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,12288,0.014577777849303352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,65536,0.061321775118509926
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,51200,0.04520977867974175
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,10240,0.012068444656001197
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,51200,0.050676445166269936
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,16384,0.01884355478816562
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,8192,0.010434666441546546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,12288,0.015156444576051501
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,16384,0.02120622164673275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,7168,0.009700444837411245
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,12288,0.017690666847758822
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,10240,0.013298666311634911
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,6144,0.009023110899660323
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,8192,0.010838222172525195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,10240,0.015308444698651632
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,5120,0.007992000215583378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,7168,0.010319111247857412
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,8192,0.01349422252840466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,4096,0.007302222152551015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,6144,0.009697777529557547
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,7168,0.012415111064910889
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,3584,0.006670222100284364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,6144,0.009992000129487779
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,5120,0.008812444077597724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,3072,0.007888889147175683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,5120,0.00903200027015474
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,3072,768,0.005250666704442766
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,4096,0.007889777421951294
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,2560,0.007309333317809635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,3072,0.007585778004593319
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,3584,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,2048,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,2560,0.007314666277832455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,3072,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,1536,0.005949333310127258
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,2560,0.006631999793979857
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,2048,0.0069751110341813825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,1024,0.0042257776690853965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,1536,0.006640000061856375
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,768,0.0038986665507157645
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,1536,0.005597333527273602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,1024,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,512,0.003141333245568805
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,1024,0.0052168890833854675
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,768,0.005931555396980709
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,256,0.002516444358560774
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,512,0.005559111220969095
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,512,0.004941333499219683
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,128,0.0021671110557185277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,256,0.005573333137565189
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,256,0.004580444345871608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,64,0.0018248889181349014
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,4096,0.00832444429397583
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,128,0.00555644432703654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2560,32,0.001812444378932317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,3584,0.007928888830873702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,65536,0.04800622330771553
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,64,0.004603555632962121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,32,0.004594666676388847
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,51200,0.038423998488320246
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,65536,0.055424888928731285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,16384,0.01482488877243466
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,65536,0.05992088715235392
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,2048,0.006232888748248418
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,12288,0.011966221862369113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,51200,0.04461244410938687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,51200,0.04855466551250882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,10240,0.0107022225856781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,16384,0.020304888486862183
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,768,0.005225777626037598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,12288,0.015104000767072042
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,8192,0.009143110778596666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,10240,0.012247111234400006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,7168,0.008644444247086843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,10240,0.014714666538768344
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,8192,0.011029333704047732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,6144,0.007616000043021307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,8192,0.011700444751315646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,7168,0.010053333308961656
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,5120,0.007017778025733099
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2560,128,0.0047688889834615915
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,6144,0.009645333720578088
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,4096,0.008997333546479544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,6144,0.008683555656009251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,5120,0.008998221821255153
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,5120,0.008388444781303406
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,3584,0.00832266691658232
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,4096,0.008293333152929941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,3072,0.0062977779242727495
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,4096,0.007256000406212277
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,3584,0.007974222302436829
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,2560,0.005586666779385672
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,3584,0.007313778003056844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,3072,0.007618666523032718
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,2048,0.007351111206743453
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,2560,0.007267555428875818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,3072,0.006630222416586346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,12288,0.014180445008807711
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,1536,0.0052471111218134565
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,1024,0.0041884444653987885
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,2560,0.006286222073766921
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,2048,0.006969778074158563
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,768,0.003903999924659729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,1536,0.006625777731339137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,2048,0.005968889014588461
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,512,0.0031724443866146933
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,1024,0.006330666856633292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,1536,0.005618666609128316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,1024,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,768,0.005932444499598608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,768,0.005223999834722943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,512,0.005769777629110549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,128,0.0021706666383478376
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,7168,0.01107022249036365
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,256,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,64,0.0019075555933846368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,32,0.0018888889915413326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,256,0.00462400002612008
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,128,0.00556355549229516
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,65536,0.03603644503487481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,128,0.004923555586073133
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,64,0.004622222234805425
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,32,0.004559999952713649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,51200,0.028425776296191748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,65536,0.054767999384138316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,65536,0.05364888906478882
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,16384,0.012181333369678922
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,16384,0.018226666582955253
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,51200,0.04452800088458591
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,12288,0.009750222166379293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,16384,0.017440888616773818
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,51200,0.04273955689536201
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,10240,0.008856000171767341
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,16384,0.017484444710943434
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,12288,0.013227555486891003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,8192,0.007610666255156199
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,12288,0.014458666245142618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,7168,0.007732444339328342
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,10240,0.012739555703269111
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,8192,0.010799110763602786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,2048,256,0.002471111093958219
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,6144,0.006962666908899943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,8192,0.009887110855844285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,7168,0.009925333162148794
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,5120,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,2048,512,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,6144,0.009312888814343346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,7168,0.009278222090668148
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,4096,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,3584,0.005400889035728242
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,6144,0.008580444587601556
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,5120,0.008637333081828224
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,3072,0.005591111050711737
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,4096,0.00833155545923445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,5120,0.007999111380841997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,2560,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,3584,0.008005333443482717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,4096,0.007288888924651676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,2048,0.00554577757914861
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,3072,0.0076453329788313965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,3584,0.0069422221018208405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,1536,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,3072,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,2048,0.006968888971540663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,1024,0.004048000193304486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,1536,0.006613333192136552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,2048,0.0058986664646201665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,768,0.0032231110251612137
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,1536,0.005574222240183089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,1024,0.006235555642180973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,512,0.002858666703104973
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,1024,0.005261333452330695
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,768,0.005936000082227919
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,256,0.0021475555582178966
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,768,0.005210666606823604
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,512,0.005595555735958948
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,128,0.002171555534005165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,512,0.004883555488453971
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,256,0.0052044445441828836
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,64,0.001820444439848264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,256,0.004563555535342959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1536,32,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,128,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,128,0.004566222015354368
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,64,0.004548444516128964
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,51200,0.020282665888468426
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,32,0.004554666578769684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,2560,0.007279111279381647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,65536,0.05091199941105313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1536,2560,0.0061039999127388
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,51200,0.043805334303114146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,12288,0.011213333242469363
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,51200,0.03732444511519538
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,16384,0.015812445018026564
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,10240,0.010024000373151567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,16384,0.015835555063353646
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,10240,0.012086222569147745
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,12288,0.013457777599493662
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,8192,0.00908622228437
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,12288,0.013086222112178802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,10240,0.011743110915025076
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,7168,0.008170666793982187
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,10240,0.011415111521879831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,8192,0.010683555569913654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,6144,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,8192,0.009545777406957414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,7168,0.009671111073758867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,5120,0.006641777853171031
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,6144,0.009684444301658207
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,4096,0.0059573331640826324
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,65536,0.02474133339193132
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,5120,0.009016888837019602
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,3584,0.005252444495757421
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,5120,0.00795999997191959
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,4096,0.00793333351612091
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,4096,0.007292444507280986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,3584,0.007666666474607255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,65536,0.05430044399367439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,3584,0.0069875551594628235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,3072,0.007491555478837755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,3072,0.006614222294754452
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,2560,0.007257777783605788
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,2560,0.006294222341643439
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,2048,0.0059004442559348206
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,2048,0.0069368887278768755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,1536,0.004915555732117759
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,2048,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,7168,0.008997333546479544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,1024,0.004568888909286923
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,1536,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,1024,0.005973333285914526
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,768,0.0038355555799272326
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,3072,0.005560889012283749
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,16384,0.012887111140622033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,2560,0.004869333157936732
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,512,0.0028684443483750024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,768,0.005603555589914322
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,256,0.002483555633160803
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,768,0.004895110925038655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,512,0.005623999983072281
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,128,0.002132444539003902
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,512,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,256,0.004543999830881755
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,64,0.0018595556418100994
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,128,0.005271111097600725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,128,0.004613333278232151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,64,0.004536889079544279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,65536,0.020101333657900494
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,32,0.004877333425813251
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,6144,0.008396444221337637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,51200,0.016319110989570618
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,65536,0.05396533343527052
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,1024,1024,0.0052266667286554975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,65536,0.046394666035970054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,16384,0.012967999610635968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,51200,0.03694222370783488
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,51200,0.04348977737956577
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,12288,0.010029333333174387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,16384,0.015513777732849121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,10240,0.008325333396593729
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,16384,0.01443555619981554
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,12288,0.013339555925793119
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,256,0.00554933316177792
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,12288,0.012084444363911947
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,10240,0.011747555600272285
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,7168,0.00701422244310379
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,10240,0.01071200023094813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,8192,0.01073066641887029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,1024,32,0.001850666685236825
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,7168,0.009694221946928237
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,5120,0.0052577778697013855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,7168,0.008955555657545725
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,6144,0.009318222602208456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,4096,0.005895111295912001
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,1536,0.006647111227114995
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,6144,0.008450667063395182
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,5120,0.008628444539176093
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,5120,0.007829333345095316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,4096,0.007989333735571967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,3072,0.004909333255555895
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,4096,0.0069413334131240845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,3584,0.007977777885066139
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,3584,0.006642666541867786
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,3072,0.007280000381999546
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,3072,0.006723555425802867
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,2048,0.004223111189073986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,2560,0.006981333096822103
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,2560,0.006232000059551663
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,1536,0.004855111241340637
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,2048,0.006598222172922558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,2048,0.005951111101441913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,1024,0.004219555606444676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,6144,0.006269333263238271
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,1536,0.005256888767083486
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,8192,0.009443555441167619
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,768,0.0032044444233179092
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,1024,0.005719999886221356
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,512,0.003077333379122946
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,1024,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,768,0.005596444424655702
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,256,0.0021866667601797315
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,768,0.004925333377387789
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,512,0.005599110904667113
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,128,0.001975111042459806
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,512,0.004917333523432414
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,256,0.005407111098368962
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,64,0.001820444439848264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,256,0.004540444248252445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,32,0.0018328889790508484
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,128,0.0052391112678580815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,65536,0.028466665082507666
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,128,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,64,0.004552888787455029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,51200,0.0236453331179089
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,768,32,0.004528000122971005
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,8192,0.007594666547245449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,65536,0.048743999666637845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,16384,0.009522666533788046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,65536,0.04494844542609321
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,12288,0.0073626670572492815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,51200,0.038584887981414795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,1536,0.006244444598754247
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,16384,0.015863999724388123
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,51200,0.03633866707483927
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,16384,0.014461333552996317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,12288,0.013054221868515015
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,8192,0.005214222189452913
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,12288,0.011748444702890186
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,10240,0.012060444388124677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,3584,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,10240,0.010668444136778513
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,8192,0.010412444670995077
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,6144,0.005631999837027655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,8192,0.009479111267460717
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,7168,0.008977777428097194
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,5120,0.0052595556610160405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,6144,0.009314667019579146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,6144,0.007992888490358988
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,4096,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,5120,0.008606221940782335
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,5120,0.007617777420414819
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,3584,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,4096,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,4096,0.006792888873153263
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,3072,0.0052488889131281115
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,3584,0.007605333295133378
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,10240,0.006633777585294511
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,2560,0.005195555587609609
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,3584,0.006635555376609166
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,3072,0.007262222468852997
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,2048,0.004503111044565837
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,3072,0.006303999986913469
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,2560,0.0070364442136552595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,7168,0.005281777845488654
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,2560,0.006245333287451003
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,2048,0.00665511108107037
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,1024,0.0031822222388452957
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,2048,0.005575110928879843
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,1536,0.006285333385070165
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,768,0.0028159999185138275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,768,2560,0.004902222090297275
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,1536,0.005490666462315454
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,512,0.002530666689078013
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,1024,0.005882666756709416
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,1024,0.004926222066084544
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,768,0.005896889087226655
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,256,0.0021884443445338146
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,512,0.005554666535721884
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,768,0.004927111168702443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,128,0.0018560000591807896
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,512,0.004929777648713854
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,64,0.0018728888697094387
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,128,0.00526577772365676
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,32,0.0018195555441909367
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,128,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,64,0.004580444345871608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,65536,0.015648888217078317
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,32,0.004256000121434529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,512,1536,0.003516444315512975
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,51200,0.013799111048380533
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,16384,0.005952888892756567
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,65536,0.042456888490253024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,7168,0.009683555199040307
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,12288,0.004878222114510006
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,16384,0.015840000576443143
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,16384,0.014094221923086377
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,10240,0.005244444641802046
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,12288,0.013101333545313941
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,12288,0.011707555916574268
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,8192,0.005215999980767568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,10240,0.012048888537618848
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,10240,0.01068800025516086
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,7168,0.004875555634498596
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,8192,0.01072622256146537
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,8192,0.009323555562231276
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,6144,0.004655999855862724
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,7168,0.010018666585286459
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,7168,0.008820444345474243
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,5120,0.004247111164861255
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,6144,0.009359999663299983
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,65536,0.046542220645480685
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,6144,0.008298666940795051
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,4096,0.004344888859324985
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,5120,0.008708444734414419
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,5120,0.007325333025720384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,51200,0.037223110596338906
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,3584,0.004554666578769684
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,4096,0.007947555846638149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,3072,0.004228444563017951
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,51200,0.03390222125583225
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,2560,0.004535111288229625
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,4096,0.007158222297827403
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,3584,0.007790221936172909
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,3072,0.007321777443091075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,2048,0.004223111189073986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,3584,0.0064151109092765385
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,2560,0.006976889239417181
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,1536,0.0035217776894569397
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,2048,0.006623999940024481
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,2560,0.005664000080691443
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,256,0.00526311124364535
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,1024,0.0028355556229750314
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,1536,0.006229333165619109
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,2048,0.005937777873542573
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,512,256,0.004576888763242298
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,1536,0.0053048887186580235
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,1024,0.005965333431959152
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,512,0.002492444382773505
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,1024,0.00480000012450748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,256,0.0021439999755885866
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,768,0.004947555561860402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,512,0.005565333283609814
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,128,0.001839111041691568
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,512,0.004885333279768626
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,256,0.005593777944644292
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,64,0.001855111163523462
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,256,0.00425866660144594
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,128,0.005568888866239124
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,32,0.0017875555074877208
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,128,0.004545777622196409
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,64,0.0047084444926844705
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,32,0.004882666799757215
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,51200,0.009371555513805812
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,65536,0.04124888777732849
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,65536,0.04585244589381748
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,256,3072,0.006279110908508301
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,16384,0.005592888842026393
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,51200,0.03727111220359802
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,12288,0.004864888886610667
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,51200,0.033598221010631986
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,16384,0.015834665960735746
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,10240,0.00488800017370118
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,16384,0.014109333356221518
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,768,0.005935110979610019
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,12288,0.01308444473478529
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,8192,0.004911111046870549
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,10240,0.012038222617573209
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,12288,0.011759111450778114
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,7168,0.005192000004980299
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,10240,0.010748444332016839
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,8192,0.01036000003417333
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,6144,0.004557333472702238
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,7168,0.010017777482668558
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,5120,0.004565333326657613
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,7168,0.009123555488056606
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,65536,0.011329777538776398
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,4096,0.0038746665749284956
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,6144,0.008310221963458592
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,5120,0.008654221892356873
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,3584,0.004576000074545543
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,5120,0.007627555893527136
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,4096,0.007979555262459649
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,3072,0.004226666771703296
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,4096,0.006575999988449945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,3584,0.007852444218264686
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,2560,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,3584,0.006923555499977536
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,3072,0.007236444287829929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,2048,0.0038577777643998465
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,256,768,0.0026062221990691293
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,3072,0.006301333506902059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,1536,0.0035253332720862497
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,2560,0.005964444329341252
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,2048,0.006576889091067844
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,1024,0.0028133332315418455
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,2048,0.0052622221410274506
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,1536,0.006573333508438534
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,768,0.0025351111673646504
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,1536,0.0052728888889153796
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,1024,0.005902222047249476
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,512,0.002455111179086897
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,8192,0.009546666509575313
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,768,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,256,0.001973333251145151
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,512,0.005585777676767773
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,6144,0.009328000247478485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,768,0.004938666605287128
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,128,0.0018542222678661346
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,64,0.0017955555684036678
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,256,0.005579555614127054
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,128,32,0.0018026667336622875
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,128,0.005584888988071018
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,256,0.004214222232500712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,65536,0.00657066661450598
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,128,0.004912000149488449
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,64,0.004581333448489507
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,16384,0.004912888838185204
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,32,0.004881777697139316
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,12288,0.004583999928500918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,65536,0.0411119990878635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,10240,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,51200,0.03326844506793552
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,8192,0.004547555413511065
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,2560,0.006937777830494775
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,16384,0.014089778065681458
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,7168,0.0045422220395671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,6144,0.004478222380081813
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,10240,0.010477333433098264
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,5120,0.004406222038798862
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,8192,0.009351111120647853
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,7168,0.008795555267069075
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,4096,0.004234666625658671
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,6144,0.007956444389290279
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,1024,0.004920888692140579
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,5120,0.007639110916190677
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,3072,0.003929777691761653
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,128,512,0.00452533322903845
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,2560,0.004208000169859992
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,3584,0.0063493334584765965
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,51200,0.005993777679072485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,3072,0.006257777826653586
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,1536,0.0035208890007601846
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,2560,0.005633777628342311
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,1024,0.0028035555862718155
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,2048,0.005255110975768831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,768,0.0024577778660588795
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,1536,0.0052355556852287715
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,512,0.0021644443687465456
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,1024,0.004879999905824661
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,12288,0.011711111499203576
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,256,0.0018515555808941524
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,128,0.0018115554832749898
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,768,0.004567999806669024
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,64,0.0017804443422291016
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,512,0.00462666650613149
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,32,0.0018293333964215384
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,256,0.00489066665371259
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,3584,0.004233777936961916
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,65536,0.00629688882165485
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,128,0.004210666649871402
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,51200,0.006234666539563074
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,16384,0.004580444345871608
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,64,0.0041786668201287585
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,12288,0.004518222063779831
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,32,0.004572444491916233
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,10240,0.004311111238267687
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,65536,0.04107466671201918
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,51200,0.03305244445800781
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,64,2048,0.0038808890514903595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,8192,0.003896000070704354
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,7168,0.004214222232500712
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,6144,0.00423200014564726
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,12288,0.011352000137170156
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,5120,0.0038168889780839286
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,10240,0.01036977767944336
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,4096,0.0038737778862317405
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,8192,0.009032888544930352
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,7168,0.008643555144468943
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,3584,0.0041777777175108595
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,3072,0.0038533334930737815
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,6144,0.007972444097201029
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,5120,0.007265778051482306
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,4096,0.0069360000391801195
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,2048,0.003922666526503033
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,3584,0.00665866666369968
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,1536,0.0034977777136696708
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,3072,0.005933333188295364
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,1024,0.0028417776856157514
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,2560,0.005544888890451855
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,768,0.0024462222225136226
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,2048,0.005588444570700328
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,512,0.0022257777551809945
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,1536,0.0052782222628593445
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,256,0.002136888810329967
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,1024,0.004871110949251386
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,128,0.0018382221460342407
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,64,0.0017831110292010838
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,512,0.004562666846646203
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,16384,0.014120888378885059
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,32,0.0018497777895794974
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,256,0.004578666554556953
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,128,0.004209777961174647
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,64,0.004603555632962121
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,32,0.004553777890072929
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,float16,1,32,2560,0.004249777644872665
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,64,4096,0.0069679998689227635
VLLM,0.17.0,NVIDIA B200,gemm,vllm_default,fp8,1,32,768,0.004554666578769684
